Hospital surge planning gets reduced to a bed count more often than any other part of emergency management, and the bed count is the least useful number in the building. This piece covers the staff, stuff and space framework and why staff binds first, how the conventional, contingency and crisis standards of care continuum actually works and who has to declare the crisis end of it, what a healthcare coalition is and what it coordinates, how decompression and patient movement create capacity that no construction project can, and why a mass casualty incident is usually a distribution failure rather than a capacity failure.

The distribution problem, stated plainly

A mass casualty incident almost never exhausts the hospital capacity of a metropolitan region. What it reliably does is deliver a large fraction of its patients to the nearest emergency department inside the first forty minutes, by private car, by police cruiser, on foot and in the arms of bystanders, which is how a community hospital a mile from the scene ends up with eighty arrivals while a Level I trauma center ten miles out has open operating rooms, a staffed trauma bay and nothing to do. The regional arithmetic looks fine on the report afterward, and it does nothing for the patient standing in the wrong lobby at 21:40.

Two parts of the trauma literature frame this well enough to be worth knowing by name. Eric Frykberg published reviews of terrorist bombing experience in the trauma journals arguing that over-triage, meaning the routing of non-critical patients into critical care resources, correlates with higher mortality among the genuinely critical, because the scarce resource gets consumed by people who did not need it. The Boston Marathon bombing of April 15, 2013 is the case usually set against that, since the scene was close to several Level I centers, EMS distributed patients across multiple hospitals rather than filling the closest, and published clinical reviews report that every patient who reached a hospital alive survived. Three people died at the scene, and injury counts for that day vary by source, commonly given as 264, with the figure shifting depending on how late-presenting patients were counted.

The opposite end shows up in the Las Vegas shooting of October 1, 2017, where a great many of the wounded moved themselves or were moved by other civilians in pickup beds and ride-share cars, and two facilities absorbed the overwhelming majority of them. The Clark County coroner’s office recorded 58 victims killed at the time, a total the county later revised to 60 after two people died of their injuries in subsequent years, and published injury counts for the event differ substantially, with early law enforcement figures in the hundreds and later tallies running above 800 depending on whether crowd-crush injuries are included. Sunrise Hospital and Medical Center has published its own account of receiving more than two hundred patients in a matter of hours. A companion article in this section covers the 2001 surge in New York and the hospitals that cleared out and then waited, so I am not going to re-argue that case here, because this piece is about the framework that the case study deliberately leaves alone.

Staff, stuff and space, and which one binds first

The staff, stuff and space formulation comes out of the surge capacity literature of the 2000s and is now standard in federal and state healthcare preparedness guidance, usually with systems added as a fourth element covering command, communications and the processes that tie the other three together. It is useful because it forces the question past the furniture. An empty hospital bed is a frame, a mattress and a headwall, and a region can produce those by the hundred out of closed units, gymnasiums and surplus storage. What makes it a functioning bed is a licensed person who can care for the patient in it, the consumables and equipment that patient requires, and a physical location with the oxygen, suction, monitoring and power that the care depends on.

Staff binds first and recovers last, because licensure, competency and currency cannot be purchased on the morning of the event. An intensive care patient on a ventilator requires a nurse with critical care competency at a ratio the unit can actually sustain across three shifts, plus respiratory therapy, plus a physician or advanced practice provider managing the vent and the drips, and you can run a two-to-one ratio for a shift or two before the error rate and the staff attrition start charging you for it. Federal pandemic planning documents from the mid-2000s used absenteeism assumptions as high as forty percent at the peak, which is the number nobody wants to plan against because it eliminates the comfortable answer of calling people in. The levers that do exist are worth knowing before you need them, since the Joint Commission’s emergency management standards permit granting disaster privileges to volunteer licensed independent practitioners under defined conditions when the emergency operations plan is activated, most states run an Emergency System for Advance Registration of Volunteer Health Professionals registry, Medical Reserve Corps units exist in many counties, and a number of states have adopted the Uniform Emergency Volunteer Health Practitioners Act to handle out-of-state licensure. Confirm how your own state handles practitioner licensure in a declared emergency with your state health department rather than assuming reciprocity.

Stuff fails in less obvious places than the supply closet. Bulk liquid oxygen systems and their vaporizers became a real constraint during the worst COVID-19 surges, because a hospital built to deliver oxygen to a normal census cannot necessarily deliver thirty or sixty liters per minute to every bed on a floor without freezing a vaporizer or dropping line pressure, and that is a plant engineering limit rather than a purchasing one. Space has its own hard edges, since negative pressure capability, oxygen-capable headwalls, telemetry coverage and sightlines for nursing are not features you add to a converted conference room over a weekend. In the years I spent on the supply and quality side of a fire department communications and logistics shop, the recurring lesson was that the item you run out of is never the item on the shortage list from last year.

Conventional, contingency and crisis: a continuum with a declaration in the middle

The three-tier language comes from the Institute of Medicine, now the National Academy of Medicine, which published a letter report in 2009 giving guidance on establishing crisis standards of care in disaster situations and followed it in 2012 with a multi-volume systems framework for catastrophic disaster response. Conventional capacity uses the usual spaces, staff and supplies, and the care delivered is consistent with daily practice. Contingency capacity uses spaces, staff and supplies that are not consistent with daily practice but are functionally equivalent, which is what you are doing when the post-anesthesia care unit holds ventilated patients, when a medical-surgical nurse takes an extra patient, or when pharmacy substitutes one agent for another on a shortage. Crisis capacity means adaptive spaces, staff and supplies that are not consistent with usual standards of care, where the hospital can no longer promise each individual the usual level of care and the operating goal shifts toward the best achievable outcome for the population of patients presenting.

Two features of this continuum get missed constantly. The first is that it is resource-specific and time-specific rather than a mode the whole hospital enters, so a facility can be in crisis for ventilators at 03:00 on Tuesday, in contingency for medical-surgical nursing all week, and entirely conventional for everything else. The second is that the crisis end is not a decision an individual clinician makes at the bedside, because the IOM framework and essentially every state plan built on it put the shift to crisis standards behind a formal process, generally a declaration by state authority, with clinical triage decisions made by a team that is deliberately separated from the physicians and nurses treating the patient, both to keep the decisions consistent across facilities and to keep the bedside clinician from carrying the allocation decision personally.

State plans vary a great deal in how they handle allocation criteria, review, appeals and liability protection, and some of those plans were revised hard between 2020 and 2022 after disability rights organizations and the HHS Office for Civil Rights raised objections to criteria that discriminated on the basis of disability or long-term life expectancy. If your agency is going to reference crisis standards in a plan, get the current version of your own state’s document from the state health department and read who declares, what protections attach, and what the clinical triage process is, because the answer is genuinely different from one state line to the next and an article is not the place to learn it.

The healthcare coalition and who coordinates what

The healthcare coalition is the structure that makes regional distribution possible, and it exists mostly because of one funding stream. The Hospital Preparedness Program began in 2002 under the Health Resources and Services Administration in the aftermath of the 2001 attacks and the anthrax mailings, moved to the federal preparedness office that is now the Administration for Strategic Preparedness and Response, and over time shifted its money from individual hospital grants to coalitions covering a defined geographic area. A coalition’s core membership is hospitals, emergency medical services, emergency management and public health, with the better ones pulling in long-term care, dialysis, home health, behavioral health, federally qualified health centers, blood suppliers, medical examiners and the dentists and veterinarians who turn out to matter in specific scenarios. The program’s appropriation has declined substantially in real terms since its first years, which is visible in how thin coalition staffing is in much of the country, and the current figure is in the HHS budget justification rather than anywhere I would quote from memory.

What a coalition actually does during an incident is information sharing, resource coordination and the multi-agency coordination function, which in practice means running the bed and resource picture, brokering transfers and supply movement between member facilities, and giving the state health department and the local EOC a single place to talk to the healthcare sector instead of calling eleven hospitals individually. The federal capability guidance published for the program in 2017 organized that work into four capabilities covering readiness, response coordination, continuity of healthcare service delivery and medical surge, and the guidance has been revised since, so get the current edition from the ASPR site before you write it into a plan. Inside a hospital, command runs on the Hospital Incident Command System, whose guidebook was substantially revised in 2014 under the California Emergency Medical Services Authority and should be checked for the current edition before anyone prints job action sheets. Above both sits Emergency Support Function 8, public health and medical services, led by the state health department at the state level and by HHS federally.

Two regulatory drivers sit underneath all of this and are worth naming because they are what get the planning done at facilities that have no other reason to do it. The CMS emergency preparedness rule, finalized in 2016 with compliance beginning in 2017 and amended in 2019 to reduce documentation burden, requires a long list of Medicare and Medicaid provider and supplier types to maintain a risk-based emergency plan, policies and procedures, a communication plan, and a training and testing program. The Joint Commission maintains emergency management standards that were restructured in a recent revision. Verify the current requirements and review cycles with CMS and your accrediting organization, since both have moved more than once.

Decompression: making beds by emptying them

The fastest route to an available bed is a discharge, and the second fastest is a transfer out, which is why decompression rather than expansion is the center of serious surge planning. Gabor Kelen and colleagues at Johns Hopkins published work on what they called reverse triage, a method for identifying inpatients at low risk of a consequential medical event over the next several days who could be discharged early to create immediate capacity, and the framing matters because it turns an unstructured hunt for discharges into a defensible clinical screen. On an ordinary weekday a large hospital discharges a substantial fraction of its census, and concentrating that work, moving the rounding earlier, and dedicating providers to discharge decisions can pull hours out of the cycle and open a meaningful number of beds in a single shift.

Cancelling elective surgery is the other large lever, and it produces more than operating rooms. It frees anesthesia providers, post-anesthesia nurses with critical care skills, monitored bays, ventilators and the pre-operative area, which is why the post-anesthesia care unit is the standard first overflow intensive care space in nearly every surge annex I have read. The accompanying cost is real and is usually understated in the plan, because cancelled cases are deferred care for patients with cancer and cardiac disease, and they are also the hospital’s margin, which is part of why the decision to stand back up gets made faster than the clinical picture supports.

Decompression fails downstream more often than it fails inside the hospital. A patient who is medically ready to leave still needs a receiving skilled nursing facility or long-term acute care bed, a ride, a prescription that a pharmacy will fill, oxygen or other durable medical equipment delivered to a home, and in some cases a family member who is reachable. During the COVID-19 surges, the binding constraint for many hospitals was not an ICU bed but a nursing home that would not accept a transfer, and discharge volume on a Saturday night in almost any hospital is close to zero because the people who authorize and arrange those discharges are not at work. If your coalition has never brought post-acute providers to the table, your decompression plan has a gap exactly where the patients leave.

Patient movement, load balancing and the scarce specialty bed

Load balancing is the operational name for fixing the distribution problem after the fact, and it requires three things that have to exist before the incident: visibility into who has what, an authority to broker the move, and transport. Visibility usually comes through a regional bed and resource platform, with EMResource and similar systems common and state systems varying widely in how current their data is, since a bed board that nobody updates between drills is a decoration. The brokering function is what ASPR promoted during the pandemic as a medical operations coordination cell, and Arizona’s state health department ran one of the better-documented examples with a statewide transfer line that load-balanced patients between facilities under a single set of rules rather than leaving each hospital to make forty phone calls. Several states have kept some version of that capability, and the right question for your region is who answers the phone and what authority they have when two hospitals both say no.

Transport is the constraint everybody forgets. Moving one ventilated intensive care patient fifty miles consumes a critical care transport crew, a specialized ambulance and often two to four hours door to door, and the region has a small number of those crews on any given night, while air medical capacity drops out entirely in weather. For large-scale movement the federal mechanism is the National Disaster Medical System, which pairs medical response teams with a patient movement function and a network of definitive care beds pledged by civilian hospitals and coordinated through Federal Coordinating Centers hosted by the Department of Veterans Affairs and the Department of Defense. It has been used in hurricane evacuations and is exercised more than it is activated, so confirm your own facility’s NDMS participation status and the identity of your Federal Coordinating Center with ASPR rather than assuming.

Specialty capacity is where the distribution problem becomes a national problem. The American Burn Association verifies burn centers at a scale of roughly sixty to seventy facilities nationwide, and its disaster planning guidance has described an expectation that a verified center can absorb on the order of fifty percent above its usual census in a disaster, which means a single large burn event consumes the regional capacity and then the capacity of several surrounding states. Pediatric intensive care is similarly regionalized, and the adult hospital that receives children from a school incident will need the pediatric referral center on the phone early for telephone consultation as much as for transfer. Two legal points belong in the plan rather than in a lawyer’s voicemail on the night, because EMTALA governs transfers and does not go away because the event is large, and waivers of selected Medicare requirements under section 1135 of the Social Security Act require both a presidential emergency or disaster declaration and a public health emergency determination by the Secretary of HHS, with the specific waivers published by CMS. Verify the current waiver scope with CMS, because what was waived in one event is not automatically waived in the next.

The front door: self-presenters, decontamination and the information desk

In chemical and radiological events the hospital is a first responder whether or not anyone planned it that way, because contaminated and worried people arrive before any field decontamination corridor exists. The Tokyo subway sarin attack of March 20, 1995 remains the reference case, with roughly five thousand people seeking medical attention across the city’s hospitals and the great majority arriving by taxi, by foot and in private vehicles rather than by ambulance, and St. Luke’s International Hospital alone received several hundred while published accounts describe secondary contamination of hospital staff from patients’ clothing and off-gassing. Death figures for that attack are reported as 12 or 13 depending on the source and the period covered, with a later death in 2020 bringing some totals to 14. The operational point is that an emergency department can be rendered unusable by the people walking through its doors, and the countermeasure is an access control plan, an outdoor triage and decontamination point, and staff who can don protective equipment in minutes rather than find the key to the cabinet it is stored in.

Access control is the quieter half of the same problem. A hospital receiving casualties from a shooting will also receive the families, the media, law enforcement conducting investigation, and in documented cases a perpetrator or an associate, which is why lockdown with controlled entry points, a staffed family reunification area physically separated from the emergency department, and a pre-identified space for law enforcement belong in the surge annex rather than in a separate security document nobody reads. Staff also have to get in, which means parking and badge access have to work under lockdown or you will have fifty nurses standing outside a locked garage.

The information burden is a capacity problem in its own right. Every patient arriving without a name generates a temporary identifier, and every family calling the main number generates a request that somebody has to answer without violating privacy rules and without guessing. Hospitals that handle this well have a pre-assigned patient tracking method, a liaison working with the local emergency management reunification effort, and a public information function that is staffed from the start, while hospitals that handle it badly discover that their charge nurses spent the first two hours of a mass casualty incident on the telephone.

What a surge exercise has to measure

A surge exercise that produces a bed number has measured almost nothing. The numbers worth generating are elapsed times and failure points: how long from notification to the first additional staffed ICU bed, how many additional patients the emergency department can physically hold with the hallway and the overflow area in use, how long to convert the PACU, how many discharges can be executed in four hours on a Sunday, how long to move the first transfer patient out the door, and how long until the decontamination team is dressed and operating. Earlier editions of the federal preparedness guidance expressed medical surge as a percentage increase in staffed beds achieved within a few hours, and whatever benchmark your coalition currently uses should be tested against the clock rather than asserted in a report.

Build the exercise so that the distribution problem actually appears. Inject patients who arrive by private vehicle rather than by ambulance, make the nearest hospital the one that receives three quarters of them, and give the coalition’s coordination function the job of fixing the imbalance in real time with the transport resources that are genuinely available at that hour. Include at least one post-acute partner so that somebody has to answer whether a skilled nursing facility will take a patient on a Saturday, and include a pediatric or burn patient so that the specialty referral path gets dialed rather than assumed.

Document the decisions, not just the outcomes. The most valuable artifact a surge exercise can produce is a written record of who authorized each move from conventional into contingency, what indicator triggered it, and what would have been needed to go further, because that record is the draft of your indicators and triggers table. The federal and IOM guidance both put indicators and triggers at the center of the framework, and a table written by people who have never watched their own hospital fill up tends to describe conditions that no charge nurse would recognize.

What to do at your agency

• Have your emergency preparedness coordinator pull the last regional bed poll your coalition ran, confirm that it reports staffed and immediately available beds broken out by unit type rather than as one aggregate number, and get the reporting categories corrected in writing if it does not.
• Ask your state health department’s healthcare preparedness office for the current state crisis standards of care plan, read who holds the declaration authority and what the clinical triage process is, and put the document and its date into your emergency operations plan reference file.
• Put contingency capacity on the agenda of the next emergency management committee meeting that already meets, and come out of it with a one-page list naming the specific contingency actions available at your facility and the role title authorized to order each one.
• Have the EMS chief and the nearest hospital’s emergency department director jointly walk the ambulance entrance and the adjacent parking area, and write a half-page plan for where an external triage and decontamination point goes, who staffs it, and how vehicle traffic is routed around it.
• Ask your coalition coordinator by name who answers the load-balancing phone in your region at 03:00, what authority that person has, and how many critical care transport units are genuinely available overnight, and record the answers rather than the policy.
• Verify this month that at least three named people besides the usual coordinator can access your regional bed and resource reporting platform with current credentials, and have one of them complete an actual report as a test.
• Invite one skilled nursing facility administrator and one home health or dialysis provider from your service area to the next coalition meeting, and ask them directly what it takes to accept a transfer on a weekend.

Takeaways

• A mass casualty incident is usually a distribution failure rather than a capacity failure, because most patients go to the nearest facility in the first forty minutes while regional capacity sits unused ten miles away.
• Staff, stuff and space is the working framework for surge, and staff binds first and recovers last because licensure, competency and currency cannot be bought on the day of the event.
• Conventional, contingency and crisis capacity come from the Institute of Medicine’s 2009 letter report and its 2012 systems framework, and the continuum is resource-specific and time-specific rather than a mode the whole hospital enters.
• The shift to crisis standards of care sits behind a formal declaration process under state authority with clinical triage decisions separated from the treating clinician, and state plans differ enough that you have to read your own.
• Healthcare coalitions exist largely because of the Hospital Preparedness Program begun in 2002, and their response value is bed and resource visibility, transfer brokering and a single point of contact for the healthcare sector.
• Decompression through early discharge, cancelled elective surgery and transfer out creates capacity faster than any expansion, and it fails most often downstream at post-acute providers, transport and weekend staffing.
• Burn and pediatric intensive care capacity is regionalized and small, with the American Burn Association verifying roughly sixty to seventy burn centers nationwide, so a single large event becomes a multi-state movement problem.
• In chemical events most casualties arrive at hospitals by private vehicle, as roughly five thousand people did across Tokyo hospitals after the 1995 sarin attack, which makes access control and external decontamination a hospital capability rather than an EMS one.