Mass Casualty Triage and Incident Command

In depth

The one-page summary above is what the PDF prints. These sections go further.

Incident command in the field and the hospital

US responses run on the Incident Command System within NIMS: one incident commander, or a unified command when agencies share authority; plain language; modular growth through sections, branches and groups; a span of control near 1:5 (NIMS; many texts give 3–7); and unity of command. The first-arriving unit establishes command, transmits a size-up (incident type, estimated patients and how many walk, hazards, entrapment, hospital distances), requests resources and starts triage. The transport officer spreads patients across hospitals by priority and specialty need and logs each tag number and destination; that log later drives reunification.

In HICS (2014, fifth edition) each position has a job action sheet; staff nurses work in the Medical Care Branch or report to the labor pool, and nobody self-deploys to the ED.

START and JumpSTART step by step

START (Newport Beach Fire and Marine Department and Hoag Hospital, 1983; updated 1996) sorts adults in under a minute each. An apneic adult gets one airway opening, with no pulse check or rescue breaths, because an adult who stays apneic is rarely salvageable while many others wait. Benson, Koenig and Schultz (1996) substituted the radial pulse for capillary refill, which is hard to judge in cold or darkness; CHEMM accepts either.

JumpSTART (Lou Romig, Miami Children's Hospital, 1995; updated 2001) covers patients who appear younger than 8 (Romig: if the victim looks like a child, use JumpSTART). A child's apnea is more often respiratory, with perfusion briefly maintained, so an apneic child with a palpable peripheral pulse gets 5 rescue breaths before being tagged dead. The rate band is 15–45 (irregular breathing is also red), perfusion is a palpable pulse in the least injured limb, and mental status uses AVPU: posturing or no response to pain is red, withdrawal is yellow. A child who cannot walk at baseline is triaged in place: any red criterion is red; otherwise yellow with significant external injury, green without.

SALT and the Model Uniform Core Criteria

Mismatched algorithms, tags and category names caused confusion when agencies met. A CDC-sponsored workgroup produced SALT in 2008, and the Model Uniform Core Criteria (Lerner, Disaster Med Public Health Prep 2011) set 24 criteria in 4 groups (general considerations, global sorting, lifesaving interventions, individual assessment) that any system must meet. MUCC is a standard, not an algorithm: SALT was built to meet it, while START and JumpSTART share some criteria but, as usually taught, not all. A lifesaving intervention is done only if it is within scope, the equipment is at hand, it takes under a minute and the provider need not stay.

In a drill with 217 assessments, SALT was correct in 81% (8% overtriage, 11% undertriage) at 28 seconds per patient (Lerner, Prehosp Emerg Care 2010). In a pediatric tabletop study of 107 prehospital professionals, SALT was among the fastest but leaned toward undertriage (16.3%); the Sacco method was most accurate (89.3%) but slowest, and no tool excelled at accuracy, speed and agreement together (Cheng, JACEP Open 2022).

Tag categories, triage accuracy and expectant care

Undertriage delays lifesaving care; overtriage spends resuscitation bays, surgeons and ambulances on the noncritical. Frykberg and Tepas (Ann Surg 1988; 3,357 casualties in 220 bombings) found mean overtriage of 59% and critical mortality rising with it (r² 0.845). A simulation (Hupert 2007) found the ratio of critical patients to staffed trauma bays and ORs mattered more than overtriage itself; both point to keeping the noncritical out of the surgical pipeline.

The first outcome study of START after a real disaster, a 2003 train crash with 148 patients, found 22 red, 68 yellow and 58 green field tags against outcome-based counts of 2, 26 and 120: 79 overtriaged, 3 undertriaged, red 100% sensitive, and red patients reaching hospital a median of over an hour earlier (Kahn, Ann Emerg Med 2009).

Expectant is a resource judgment, not a diagnosis. SALT allows palliative care when resources suffice (CHEMM), and every tag is re-triaged as resources change. Move expectant patients to a sheltered area with a clinician assigned; give opioids for pain and air hunger, keep them warm, record drugs and times on the tag, and involve chaplains and family.

Secondary triage and the hospital's first hours

The CDC blast primer warns of 'upside-down' triage: the less injured bypass EMS and reach the closest hospital first, and the most severely injured arrive later. Half of initial casualties seek care within about an hour, so doubling the first hour's arrivals roughly predicts the first wave. After the 1995 Tokyo sarin attack, with no field decontamination, St Luke's International Hospital saw 641 victims that day; 5 were critical and another 106 were admitted (Ohbu, South Med J 1997).

• One secured entrance, with decontamination outside it if contamination is possible
• A triage officer at the door re-triages every arrival, whatever the field tag
• Clear the ED, stop elective surgery, and start reverse triage on the wards and ICU
• Pre-made disaster identities linked to tag numbers; photograph unidentified patients
• Uncrossmatched group O blood and the massive transfusion protocol ready
• Separate red, yellow, green and expectant areas, each with its own team

Secondary triage repeats with fuller physiology, the injuries found and the resources actually present. Green patients still need a clinician for missed eye, ear and abdominal blast injuries. When evacuation is impossible for days, SAVE (Benson 1996) weighs benefit against resources consumed: an elderly patient with 70% TBSA burns is expectant, while a young adult with GCS 12 who needs only airway care is treated.

Surge capacity and crisis standards of care

Planning covers space, staff, supplies and systems, and US hospitals are expected to manage up to 96 hours without outside aid (Kelen 2009). Reverse triage supplies most new beds: in three hospitals, 44% of 3,491 inpatients needed no critical intervention over the next 4 days, reverse triage gave 50–59% of surge beds, and most capacity appeared within 24–48 hours (Kelen, Disaster Med Public Health Prep 2009). ICU contingency steps: transfer the most stable patients with a full handoff, open the PACU, use anesthesia machines and transport ventilators, and let ICU nurses lead teams.

The IOM (2009) defines crisis standards of care as a substantial change in usual operations and the level of care possible, made necessary by a pervasive or catastrophic disaster and formally declared by a state government. Its ethical base is fairness, the duty to care, the duty to steward resources, transparency, consistency, proportionality and accountability. A triage officer or team applies written criteria so bedside clinicians do not allocate, which limits bias and moral distress. The Ontario protocol (Christian, CMAJ 2006) used SOFA-based criteria, reassessed at 48 and 120 hours, and coded blue the patients who should not receive critical care.

COVID-19 exposed the weak points: SOFA discriminated poorly in ventilated COVID-19 pneumonia (AUROC 0.59, below age alone at 0.66; Raschke, JAMA 2021), HHS's Office for Civil Rights warned in 2020 against denying care on the basis of disability or age, and withdrawing a ventilator for reallocation remains contested.

CBRNE: decontamination, nerve agents and radiation

At the hospital, OSHA 3249 names the decontamination zone and post-decontamination zone and sets minimum first-receiver PPE in the former: a powered air-purifying respirator with an assigned protection factor of 1,000, a chemical-resistant garment, double gloves and chemical-protective boots. Disrobing removes most contamination (about 75–90% in OSHA 3249; up to 90% of radioactive material in REMM). PRISM Volume 2 adds dry decontamination, a 15–30 second ladder-pipe rinse, and if needed technical decontamination: 60–90 seconds with water at 95–104 °F (35–40 °C) and mild detergent. Airway, hemorrhage control and antidote autoinjectors should not wait for decon.

Nerve agents inhibit acetylcholinesterase. Vapor acts in seconds to minutes: miosis, rhinorrhea and chest tightness, then bronchorrhea, bronchospasm, fasciculations, seizures and apnea. Liquid on skin can take up to 18 hours to cause systemic effects (Merck 2026). In Tokyo, over 20% of St Luke's staff had symptoms of secondary exposure (Ohbu 1997). Intubate with rocuronium; succinylcholine's effect is prolonged.

In radiation incidents, stabilization comes before decontamination (REMM); a dirty bomb injures mainly by blast. An exposed patient is not radioactive; a contaminated one carries material but is low risk to staff in standard precautions with dosimeters. Remove clothing, then shrapnel, irrigate wounds and orifices, then wash skin gently with tepid water and mild soap, aiming for no more than twice background (REMM).

Earlier vomiting means a higher dose: about 2–6 hours after 1–2 Gy, 1–2 hours after 2–6 Gy, and under an hour above 6 Gy (Merck); anxiety confounds it, so serial lymphocyte counts refine the estimate. Combined radiation and trauma has a worse prognosis; REMM puts the window for surgery before cytopenias at 24–36 hours, older military guidance at 48. Potassium iodide (adult 130 mg) protects only the thyroid from radioactive iodine; Prussian blue binds cesium and thallium, and DTPA chelates plutonium, americium and curium.

Blast injuries and active shooter events

Blast lung, the most common fatal injury among initial survivors, usually shows on arrival but has appeared up to 48 hours later: dyspnea, cough, hemoptysis, a 'butterfly' chest X-ray and the triad of apnea, bradycardia and hypotension (CDC). Manage it like pulmonary contusion, and watch for pneumothorax and arterial air embolism, which can mimic stroke, MI or acute abdomen. CDC advises a chest X-ray for everyone exposed and a chest tube before general anesthesia or air transport if blast lung is suspected.

Isolated TM rupture is not a marker of serious injury; traumatic amputation is a marker of multisystem injury. Abdominal injury can stay silent, and up to 10% of survivors have significant eye injuries (CDC). With similar devices, mortality was 49% on buses and 7.8% in the open air (Leibovici, J Trauma 1996). Wounded survivors without reliable hepatitis B vaccination get vaccine within 24 hours if possible, plus tetanus prophylaxis; HIV PEP is generally not warranted (MMWR 2008).

The Hartford Consensus (ACS, 2013) set the THREAT sequence: threat suppression, hemorrhage control, rapid extrication, assessment, transport. Police-escorted rescue task forces control bleeding and open airways in the warm zone. Among 139 deaths in 12 public mass shootings, only 7% had potentially survivable wounds, mostly chest, and none died of extremity hemorrhage (Smith, J Trauma 2016), so plans must go beyond tourniquets to chest seals, decompression and fast transport. Of 154 US hospital shootings in 2000–2011, the ED area was the most common site, at 29% (Kelen, Ann Emerg Med 2012). The 2014 federal guide for health care facilities applies Run, Hide, Fight and says no one can or should be told they must stay or leave.

Reunification, psychological first aid and staff resilience

Families converge on the closest hospitals. The HICS Patient Family Assistance Branch runs a family assistance center away from the ED and the media, staffed by social work, chaplains and behavioral health. Reunification depends on tracking: transport log, tag number, disaster identity, and photographs of unidentified patients. Deaths are told in person, privately, in plain words such as 'died'.

Unaccompanied children go to a supervised pediatric safe area and are released only to a verified guardian. In presidentially declared disasters FEMA can activate NCMEC's National Emergency Child Locator Center, and hospitals can report separated children up to age 18 to NCMEC's Unaccompanied Minors Disaster Registry. HIPAA stays in force but allows sharing with family involved in care and with disaster relief agencies; a section 1135 waiver can lift some sanctions, such as for speaking with family without the patient's agreement, for up to 72 hours after the hospital activates its disaster protocol.

Psychological first aid (NCTSN and National Center for PTSD, 2006) has eight core actions: contact and engagement, safety and comfort, stabilization, information gathering, practical assistance, connection with social supports, information on coping, and linkage with services. It builds on Hobfoll's five elements (2007): safety, calming, efficacy, connectedness and hope. Single-session psychological debriefing is not recommended: a Cochrane review (Rose 2002) found no benefit and possible harm, and WHO (2013) advises against it.

Plan staff endurance like any resource: set shift lengths, rest and rehabilitation, food and fluids, a buddy system, and peer support afterward. Stress reactions are expected and usually fade; symptoms beyond a month suggest PTSD and need referral. Written allocation criteria and ethics support ease the moral distress of crisis care.

Exam pearls

• First unit on scene: command, size-up, resources, then triage; it does not treat
• START red: RR over 30, no radial pulse or refill over 2 s, or not obeying. JumpSTART red: RR under 15 or over 45, no peripheral pulse, or an inappropriate or absent response to pain; an apneic child with a pulse gets 5 breaths
• SALT: still first, wavers second, walkers last; 2 breaths for an apneic child; gray expectant separate from black dead; meets MUCC
• Reverse triage: after lightning, treat the apparently dead first; in hospitals, discharge stable inpatients to free beds
• The least injured reach hospitals first; double the first hour's arrivals to estimate the first wave
• Chemical: decontaminate before the ED door; give atropine until secretions dry; avoid succinylcholine
• Radiation: stabilize first; contaminated patients pose little risk to staff; potassium iodide protects only against radioactive iodine
• Blast: isolated TM rupture is not a marker; blast lung can appear up to 48 hours later