The Medical Screening of Patients to Determine Their Priority of Need and the Proper Place for Treatment Is Called Triage

The medical screening of patients to determine their priority of need and the proper place for treatment is called triage. The word covers both halves of the job: ranking who is seen first and routing each patient to the right setting, whether that's a resuscitation bay, a fast-track clinic, a ward bed, or a discharge home with follow-up. Triage is a sorting decision made with incomplete information, usually within minutes of arrival, and it is performed by a trained clinician rather than at the front desk. The boundary matters: triage assigns urgency and location, not a diagnosis. It gets revisited as new facts arrive, and a wrong first call can be corrected without anyone losing their place in line.

The term comes from the French verb trier, to sort, and entered medical use through battlefield casualty clearing in the Napoleonic era and again in World War I. Modern civilian triage took shape in the 1960s as emergency departments filled with walk-in patients who had no family doctor. According to the Emergency Nurses Association, which publishes the Emergency Severity Index handbook now in its fifth edition, triage is a brief assessment designed to answer two questions: how sick is this patient, and where should this patient go next. Everything else in this article builds on those two questions.

A realistic example: a 58-year-old man walks into an emergency department (ED) in a mid-sized U.S. city complaining of chest pressure that started 40 minutes ago. The triage nurse notes his skin is pale and clammy, gets an ECG within 10 minutes, and assigns him to the highest acuity tier. He bypasses the waiting room entirely. A second patient, a 24-year-old with a sprained ankle and normal vital signs, is sorted to a lower tier and routed to a fast-track area staffed by a nurse practitioner. Both were "screened" at the same desk within five minutes of each other; their paths diverged because of the sorting decision, not the order of arrival.

Triage is not the same as diagnosis, and it is not the same as registration. Registration collects insurance and demographics. Diagnosis is the physician's job after workup. Triage sits between them: it decides how fast and where. A common misuse is treating the triage score as a permanent label. In practice, a triage category is a snapshot that expires the moment a patient's condition changes, and most EDs formally re-assess patients who wait more than a set interval. The scope note for this article: triage systems differ by country, hospital, and even by unit within a hospital, so the specific scales named below should be confirmed against your local protocol.

Five-Level Scales Replaced Three-Level Sorting Because Three Tiers Under-Discriminated

Most English-speaking emergency systems moved to five-level triage because three-level scales pushed too many patients into the middle category. With "emergency, urgent, non-urgent," a broken femur and a mild asthma flare could land in the same tier, which made the middle tier useless for allocating beds. Five levels spread that mass across three intermediate tiers and let departments measure how often the sickest patients actually got seen fast.

The Emergency Severity Index (ESI) is the dominant five-level tool in U.S. emergency departments. It was developed by nurses Richard Wuerz and David Eitel and is now published by the Emergency Nurses Association. ESI level 1 means immediate life-saving intervention is needed. Level 2 means high risk or a new confusion, lethargy, or severe pain. Levels 3, 4, and 5 are assigned by counting the resources a patient is likely to need, from two or more resources down to none respectively. That resource-counting step is the part most people outside emergency nursing have never heard of, and it is what makes ESI reproducible.

The United Kingdom uses the Manchester Triage System or the Royal College of Physicians' standards, and Australia uses the Australasian Triage Scale (ATS), which runs from category 1 (immediate) to category 5 (non-urgent, seen within 120 minutes). Canada uses the Canadian Triage and Acuity Scale (CTAS), a five-level scale with target physician assessment times that range from immediate at level I to 120 minutes at level V for emergency-department patients. The scales are not interchangeable: a "level 3" in one system does not mean the same clinical picture as a "category 3" in another.

System Levels Worst tier Best tier Setting
ESI (U.S.) 5 1 = immediate intervention 5 = no resources Emergency department
ATS (Australia) 5 1 = immediate 5 = within 120 min Emergency department
CTAS (Canada) 5 I = resuscitation V = 120 min target Emergency department
Manchester (U.K.) 5 Red = immediate Blue = non-urgent Emergency department
NATO / military 4 P1 immediate P4 expectant Field and casualty clearing

A failure case worth knowing: in mass-casualty triage, the "expectant" category exists specifically for patients whose injuries are so severe that treatment would consume resources needed to save multiple others. Civilian five-level scales have no equivalent, which is one reason a hospital cannot simply import a battlefield sort tool during a disaster. The 2017 Las Vegas shooting, where more than 500 patients arrived at area hospitals within hours, forced receiving facilities to blend routine ESI triage with casualty triage principles on the fly.

Emergency Departments Run a Ten-Minute Sorting Sequence Before Any Physician Sees the Patient

The sorting decision is a sequence, and the sequence matters more than the paperwork. Here is how a typical ESI-based triage encounter unfolds, with the parts that experienced nurses do not skip.

1: Immediate visual and verbal check at the doorway. The nurse looks for airway compromise, uncontrolled bleeding, altered mental status, or a patient who cannot speak in full sentences. Any of these triggers an override to ESI level 1 without further data collection. 2: Focused history and first set of vital signs. Blood pressure, heart rate, respiratory rate, oxygen saturation, temperature, and pain score. Abnormal values get flagged immediately to the charge nurse rather than filed for later. 3: Risk-factor screen for high-acuity presentations. Chest pain with cardiac risk factors, suspected stroke within the treatment window, sepsis screening criteria, and pregnancy-related complaints each carry a protocol that forces a higher tier regardless of how well the patient looks. 4: Resource prediction for ESI levels 3 through 5. The nurse estimates how many distinct resources the visit will consume: lab work, ECG, imaging, IV fluids, specialty consultation, or a procedure. Two or more resources means level 3; one means level 4; none means level 5. 5: Assignment of tier and physical destination. The tier goes into the record, and the destination β€” resuscitation room, monitored bed, general bed, fast track, or waiting room with re-check interval β€” is chosen in the same step. 6: Re-assessment at set intervals. Patients who remain in the waiting room are re-checked at intervals the department sets, typically every 30 to 60 minutes for mid-tier patients, with any deterioration triggering a re-triage and a higher tier.

Failure handling: if a patient's condition changes while waiting, the correct action is to re-triage immediately, not to wait for the scheduled re-check. Most departments have a "call-out" protocol that lets any staff member escalate a waiting patient to a clinician without going through the charge nurse. If the department is at capacity and an ESI-2 patient cannot be moved out of the waiting room, that becomes a documented crowding event, and in U.S. hospitals it triggers the same escalation pathway used for boarding admissions.

Resource-Counting Under ESI Separates Level 3 From Level 4 More Than Vital Signs Do

The resource count is the least intuitive part of ESI and the part most likely to be misapplied. It is a prediction, not a record: the nurse estimates what the visit will require before the physician has ordered anything. A patient with abdominal pain who will need labs, an ultrasound, and an IV gets counted as three resources and lands at level 3. A patient with a simple laceration who needs only wound closure lands at level 4.

According to the Emergency Nurses Association's ESI handbook, fifth edition, the counted resources are lab tests, ECG, imaging, IV fluids, specialty consultation, and simple procedures. Excluded from the count are history and physical exam, point-of-care testing, oral or topical medications, tetanus shots, prescriptions, and simple wound care that doesn't require sedation. That exclusion list is where new nurses make mistakes. A patient who gets a prescription and a bandage change is still a level 5 by resource count, even though it feels like a lot of activity.

A worked comparison: two patients arrive at the same minute. Patient A has a suspected kidney stone, needs urinalysis, a CT scan, and IV pain medication, so the resource count is three and the tier is 3. Patient B has a fingertip laceration that needs two sutures and a dressing, so the count is one and the tier is 4. If both had been placed in a three-level system, both would likely have been labelled "urgent," which tells the charge nurse nothing about which one to move first. The five-level split carries that information implicitly.

The operational consequence is measurable. A department that consistently mis-sorts level 4 as level 3 will pull beds away from genuinely sick patients and lengthen the queue for imaging. Some hospitals audit triage accuracy by comparing the assigned tier with the patient's final disposition and diagnosis, and a common finding in published quality reviews is that under-triage of level 2 patients is more dangerous than over-triage of level 4 patients. Over-triage costs time; under-triage costs outcomes.

Australia's ATS Attaches a Target Time to Every Category, Which U.S. ESI Does Not

The clearest structural difference between national scales is whether each level carries a clock. The Australasian Triage Scale does. Category 1 requires immediate treatment, category 2 requires treatment within 10 minutes, category 3 within 30 minutes, category 4 within 60 minutes, and category 5 within 120 minutes. Those targets are part of the standard, not internal targets set by each hospital, and Australian EDs report performance against them.

U.S. ESI does not attach a time target to each level in the same way. ESI defines acuity and resource use; the wait-time targets are set by departmental policy and by accreditation expectations rather than embedded in the scale itself. That means two U.S. hospitals using ESI identically can still have very different waiting-room behavior for a given tier, because the operational commitment differs even when the sorting logic doesn't.

Canada's CTAS sits between the two: it is a five-level scale that specifies a target physician assessment time for each level, ranging from immediate at level I to 120 minutes at level V, and Canadian EDs report against those targets through the Canadian Institute for Health Information. The national reporting structure is the difference-maker. A scale with time targets produces accountability data that a scale without them cannot.

Region also changes the front door. In the U.K., triage may happen in the emergency department, but the same sorting logic also runs through NHS 111 telephone assessment and through GP out-of-hours services, which route patients before they ever reach a hospital. In rural Australia, a retrieval service may re-triage a patient a second time before air transport. The scale is national; the setting is not.

Fast Track, Observation, and Resus Are Placement Decisions That Triage Makes Explicitly

The second half of the definition β€” the proper place for treatment β€” is not a formality. Placement is where triage earns its keep, because the same acuity can require very different destinations depending on what the department has running that day. An ESI-3 patient with abdominal pain might go to a monitored bed if the department suspects cardiac involvement and to a general bed if it doesn't.

Fast track is a common destination for ESI-4 and ESI-5 patients, usually staffed by advanced practice providers, and it exists specifically because those patients would otherwise consume full ED bed time for minor problems. Observation units are a different placement: they hold patients who need serial re-assessment, typically over 12 to 24 hours, before a disposition decision. Resuscitation rooms are reserved for ESI-1 and the highest-risk ESI-2 patients, and they are budgeted as fixed capacity, which is why a single ESI-1 admission can push the entire department's flow for hours.

Placement is also where triage intersects with bed management. In a crowded department, a triage nurse may assign a correct tier and still have no appropriate bed to send the patient to. That mismatch is the operational definition of ED crowding, and it's why some systems run "triaged out" or "provider-in-triage" models where a physician or APP sees the patient at the triage desk itself. Those models change the workflow without changing the underlying sorting logic.

A failure scenario to plan for: a patient triaged to fast track whose condition deteriorates after placement. Fast-track areas are typically staffed with fewer nurses per patient than the main department, so the re-assessment interval matters more there, not less. Departments that place low-acuity patients in fast track without a clear escalation route back to the main department have created a structural blind spot.

Pediatric and Obstetric Triage Use Age-Adjusted Normal Ranges That Adult Scales Do Not

Adult triage thresholds do not transfer cleanly to children. A heart rate of 130 is alarming in a 40-year-old and unremarkable in a febrile two-year-old. That single fact forces pediatric triage to use age-based vital sign ranges, and it is why pediatric emergency departments often use a modified ESI with pediatric-specific vital sign tables rather than the adult version.

The Pediatric Emergency Care Applied Research Network (PECARN) tools are an example of pediatric-specific decision support that interacts with triage: they help clinicians decide which head-injured children need imaging, which is exactly the resource-count question ESI asks. A child triaged as ESI-3 because the nurse predicts a CT scan may, after PECARN criteria are applied, need no imaging at all β€” and then the resource count and the tier both change. This is a live example of triage being revisited rather than fixed.

Obstetric triage has its own logic again. A pregnant patient with abdominal pain and bleeding is triaged to an obstetric unit rather than a general ED in many systems, and the sorting criteria include fetal status and gestational age alongside maternal vital signs. The placement decision is the whole point: the same maternal acuity can require a labor and delivery bed, an operating room on standby, or a general ED bay depending on gestational age and presentation.

Failure case: a pregnant patient at 34 weeks presents to a general ED with a headache and elevated blood pressure. A triage nurse using adult ESI may assign level 2 on the basis of the blood pressure alone, which is correct, but the placement decision should route her to an obstetric setting capable of fetal monitoring. Departments that lack that pathway need an explicit transfer protocol, and that protocol should be written before the patient arrives, not improvised at 3 a.m.

Telephone and Digital Triage Sort Patients Without Physical Examination, So the Rules Tighten

Not all triage happens in a doorway. NHS 111 in England, Healthdirect in Australia, and similar nurse-led telephone services run a structured triage interview and then route the caller to self-care, a GP appointment, an urgent care centre, or an ambulance. The sorting logic is the same two questions β€” how urgent, and where β€” but the information is thinner, so the rules compensate.

The compensation is algorithmic. Telephone triage systems use decision-support software with mandatory questions and hard stops that prevent the nurse from skipping a red-flag screen. A caller reporting chest pain with radiation to the arm cannot be routed to self-care regardless of how well they sound on the phone. That software-enforced safety net is the direct analogue of the doorway visual check in a physical ED.

Digital symptom checkers occupy a weaker position. They are consumer tools, not clinical triage systems, and their safety profile depends on the quality of the underlying rules. A well-known limitation is that self-reported symptoms miss what a nurse would notice in person: work of breathing, skin color, gait, and the patient's own affect. Treating a symptom-checker output as a triage category is a category error, even when the app uses clinical language.

Failure handling in telephone triage is protocol-driven. If the nurse cannot complete the red-flag screen because the caller is too breathless to speak in sentences, the default is to dispatch an ambulance rather than to keep interviewing. A similar default exists for callers who report symptoms in a child under a set age: the threshold for escalation drops, because young children deteriorate faster than adults with the same complaint.

Triage Accuracy Is Audited Against Final Diagnosis, and Under-Triage Is the Error That Matters

Hospitals measure triage quality by comparing the assigned tier with what actually happened. If a patient triaged as level 3 turns out to have had a myocardial infarction, that is an under-triage event, and it goes into a review process. If a patient triaged as level 2 turns out to have had a muscle strain, that is over-triage, which costs throughput but not safety.

The asymmetry is the point. Published quality reviews of emergency triage commonly report that under-triage of high-acuity patients is the error class associated with harm, while over-triage is associated with delay. A department that over-triages aggressively will look inefficient on flow metrics and safer on outcome metrics; a department that under-triages to protect flow will look efficient and will eventually produce a bad outcome. The right balance depends on the mismatch between demand and capacity, which is why triage accuracy is a management question as well as a clinical one.

Audit methods vary. Some departments review a sample of level 2 patients monthly and compare the triage note with the final diagnosis. Others track the "door-to-doctor" interval for each tier and look for outliers. A more robust approach, used in some academic centers, is to review every patient who was upgraded from a lower tier to a higher tier after initial triage, since each upgrade is a captured under-triage event.

One honest limitation: triage audit is retrospective and depends on documentation quality. A nurse who under-triages but documents a vague complaint that later "changes" may not generate the upgrade signal the audit is looking for. Systems that pair triage audit with a structured complaint vocabulary get cleaner signals, because the complaint at triage can be compared directly with the discharge diagnosis.

Mass-Casualty Triage Uses Four Categories and Reverses the Usual Rule About the Sickest First

In a mass-casualty event, the sorting priority inverts. In routine ED triage, the sickest patient is seen first, always. In a mass-casualty scene with more patients than resources, the patient with the most severe injuries may be categorized as expectant β€” cared for palliatively rather than treated definitively β€” because treating them would consume the resources needed to save several others. That inversion is the hardest part of casualty triage to teach, and it is the part that civilian clinicians are least prepared for.

The North Atlantic Treaty Organization (NATO) system uses four categories: P1 immediate, P2 delayed, P3 minimal, and P4 expectant. The U.S. system often uses a color scheme with the same logic: red for immediate, yellow for delayed, green for minimal, black for expectant. The categorization is done at the scene, usually by the first arriving medical responder, and then re-done at the casualty clearing point or hospital door.

A concrete comparison: in routine ED triage, a patient with 90% total body surface area burns and inhalation injury would be an ESI-1 and would receive immediate maximal intervention. In a mass-casualty scene with 40 other casualties and limited transport, that same patient may be categorized expectant, because the intervention needed exceeds the resources available and because the survival probability is very low compared with the other casualties. The decision is documented and revisited if resources change.

The boundary between routine and casualty triage needs a trigger. Without a declared incident, a single hospital cannot unilaterally switch a patient from treatment to expectant care. The switch requires an incident declaration from the relevant emergency management authority, and the protocols for that declaration differ by jurisdiction and should be checked with local emergency management. This is a place where the distinction between sorting and rationing becomes legally and ethically explicit.

FAQ

Is triage the same as an emergency department's initial assessment?

No β€” in most departments the initial assessment and the triage decision are separate steps that may happen minutes apart. Triage assigns tier and destination; the initial physician or advanced practice provider assessment produces the working diagnosis and treatment plan. The two get conflated because they often happen in the same bay within the first 15 minutes.

What happens if a patient is triaged too low and then deteriorates?

The department's re-assessment protocol is meant to catch it, and most EDs also have a standing escalation route that lets any staff member raise concern without going through the charge nurse. The documented re-triage and the upgrade are both expected; a department that treats the original tier as fixed is not following the standard.

Can a patient refuse a lower triage category and demand to be seen first?

Patients can ask for a re-check, but the tier is a clinical judgment and not a queue ticket that can be upgraded on request. A request should be treated as a re-assessment trigger, and a nurse who finds no change documents the conversation and keeps the original tier.

Does a higher triage number mean longer wait times everywhere?

Not necessarily, because ESI has no embedded time targets while Australia's ATS and Canada's CTAS attach explicit targets to each category. In an ESI hospital, the wait for a level 4 patient depends on departmental policy and current volume, which is why two U.S. hospitals with identical ESI training can show very different wait patterns.

How long does triage take in a busy emergency department?

The clinical portion usually takes 3 to 10 minutes for a straightforward presentation, with longer interviews for complex or high-acuity patients. The variable is not the interview but the queue: a triage nurse may complete the assessment in five minutes and then wait for a bed for 20, which is why triage time and door-to-provider time are reported separately.

Is telephone triage legally the same as in-person triage?

No β€” it is a distinct service with its own standards and its own liability framework, and the nurse's decision support is software-enforced precisely because the physical exam is missing. The output routes a caller to a setting; it does not replace the in-person triage that happens when the patient arrives.

What is the difference between triage and "streaming" in emergency departments?

Streaming is the operational routing of a patient to a care area after triage, and it uses the triage tier as input. A department can triage correctly and still stream badly β€” for example, sending a level 3 patient to fast track because the main department is full β€” and that distinction is useful when diagnosing flow problems.

Can triage be done by a paramedic at the scene?

Yes, and it routinely is. Field triage by emergency medical services uses a separate protocol, such as the CDC Field Triage Guidelines for trauma, which routes patients to the closest appropriate facility or a trauma center. Those guidelines are the pre-hospital analogue of hospital triage and use different criteria.

This article is general educational information about triage systems and does not constitute a diagnosis, treatment plan, or clinical recommendation for any individual patient. Triage scales, target times, and routing protocols vary by country, region, and hospital, so any specific decision should be confirmed with a locally qualified clinician or the relevant emergency service. This is not a substitute for emergency assessment. If you or someone near you is in immediate danger or at risk of self-harm, contact local emergency services or a local crisis line right away.