NEXUS Criteria: The 5 Low-Risk C-Spine Clearance Rules

The NEXUS Criteria are a five-item clinical decision rule used to determine which blunt-trauma patients can safely skip cervical-spine imaging entirely. They come out of the National Emergency X-Radiography Utilization Study (NEXUS), a large multicenter effort to reduce the routine, low-yield cervical-spine radiography that had become standard practice in emergency departments. For hospital patient-safety, quality and ED leadership, NEXUS matters less as a memorized checklist and more as a governance question: it is one of the primary evidence-based instruments a department can point to when it needs a documented, auditable basis for not imaging a trauma patient, alongside the related question of when it should image instead.

Where the NEXUS Criteria come from

The criteria were derived and validated by Hoffman and colleagues for the National Emergency X-Radiography Utilization Study Group, published in the New England Journal of Medicine in 2000 (“Validity of a Set of Clinical Criteria to Rule Out Injury to the Cervical Spine in Patients with Blunt Trauma,” NEJM 2000;343:94-99). The study enrolled more than 34,000 blunt-trauma patients across 21 U.S. emergency departments who underwent cervical-spine radiography, and tested whether a simple five-criterion checklist could identify, in advance, which patients had a clinically negligible probability of injury. Applied against roughly 810 patients with a confirmed cervical-spine injury in that cohort, the rule performed with approximately 99% sensitivity and a reported specificity around 13% — high enough sensitivity to be usable as a rule-out instrument, with the low specificity meaning most patients still flagged for imaging by the rule did not actually have an injury (an expected and accepted trade-off for a rule-out tool built to minimize missed fractures).

The five low-risk criteria

A patient is considered clinically clear of cervical-spine injury under NEXUS — and imaging can reasonably be withheld — only if all five of the following are true. A single unmet criterion takes the patient out of the low-risk group and imaging is indicated:

• No posterior midline cervical tenderness. No tenderness on palpation of the posterior midline of the neck, from the nuchal ridge to the first thoracic vertebra.
• No focal neurologic deficit. No new motor or sensory finding referable to the cervical spine on exam.
• Normal level of alertness. The patient is oriented and able to participate reliably in the history and physical exam — a Glasgow Coma Scale of 15, no significant head injury clouding the exam, and no other cause of altered mentation.
• No evidence of intoxication. No signs of alcohol or drug intoxication that would make the exam unreliable.
• No painful, distracting injury. No injury elsewhere (a long-bone fracture, a significant visceral injury, a large laceration) severe enough that it could mask the pain of a cervical-spine injury on exam.

Notice that three of the five criteria (alertness, intoxication, distracting injury) exist specifically to protect the reliability of the other two — a tender-to-palpation exam and a neurologic exam are only informative if the patient can accurately report and localize pain and sensation. That structural logic is worth stating explicitly in any order set or documentation template built around NEXUS, since it’s the part most likely to get compressed into a checkbox and lose its meaning.

How it’s applied in practice

NEXUS is a rule-out instrument, not a diagnostic one. A patient who meets all five criteria is classified low-risk and imaging can be withheld; a patient who fails even one criterion is not thereby diagnosed with an injury — they simply fall outside the population the rule was validated to clear, and standard imaging workup proceeds. Documentation should record each of the five criteria individually (not just a summary “NEXUS negative” note), both because that’s what a chart audit or quality review actually needs to verify appropriateness, and because a vague summary conclusion is exactly the kind of shortcut that erodes a decision rule’s real-world accuracy over time.

NEXUS vs. the Canadian C-Spine Rule

NEXUS is frequently discussed alongside the Canadian C-Spine Rule (CCR), a separate, later cervical-spine clearance instrument. Both are real, validated tools used for the same general purpose — deciding which blunt-trauma patients need cervical imaging — but they are structured differently and were not derived to be interchangeable:

• Structure. NEXUS is a flat, five-item checklist applied uniformly. The CCR is a tiered, branching rule: it first checks for high-risk factors that mandate imaging outright (such as age 65 or older, or a dangerous mechanism of injury), then checks for low-risk factors that permit safe assessment of range of motion, then has the patient actively rotate their neck to confirm mobility before clearing them.
• Population. The CCR was derived and validated specifically in alert (GCS 15), hemodynamically stable trauma patients, and is not intended for use outside that population. NEXUS’s derivation cohort was broader, covering blunt-trauma patients undergoing cervical imaging more generally.
• Head-to-head evidence. The two rules were directly compared in a prospective cohort of 8,283 alert, stable trauma patients across nine Canadian emergency departments (Stiell et al., NEJM 2003;349:2510-2518), of whom 169 (2.0%) had a clinically important cervical-spine injury. In that study, the CCR was more sensitive (99.4% vs. 90.7%) and more specific (45.1% vs. 36.8%) than NEXUS, and the authors concluded that, for the alert-and-stable population both rules target, the CCR would have resulted in less radiography with fewer missed injuries.

That result is real and worth citing accurately, but it doesn’t make NEXUS redundant. The comparison applies to the specific population the CCR is designed for; NEXUS remains the more widely taught and more broadly applied instrument in many U.S. departments, and its simpler, non-branching structure is easier to embed in a single documentation prompt. Departments should pick one rule and apply it consistently — and document which one — rather than mixing criteria from both.

Role in reducing unnecessary imaging

The operational reason NEXUS exists is imaging stewardship: before large validated decision rules like NEXUS and the CCR were in routine use, cervical-spine radiography after blunt trauma was ordered far more liberally than the actual injury rate justified, exposing patients to radiation and cost with little diagnostic yield. A well-documented, criteria-based rule-out gives a department three things a purely clinical “gestalt” decision doesn’t: a consistent standard across shifts and clinicians, an auditable record for quality review of why imaging was or wasn’t ordered, and a defensible basis for the small number of cases where a chart is later reviewed after an adverse outcome. This is the same governance logic behind other rule-out instruments used in the same clinical space, such as the Ottawa Ankle Rules for ankle and foot imaging.

Limitations and clinical judgment

NEXUS was derived and validated for blunt trauma; it is not intended for penetrating neck or spine trauma, which follows a different evaluation pathway. It also depends entirely on a reliable exam — which is precisely what three of its five criteria are checking for — so it cannot be applied, even in spirit, to a patient whose alertness, sobriety, or distracting-injury status is genuinely in question; those patients are correctly excluded by the rule itself, not exceptions to it. As with any clinical decision instrument, NEXUS is meant to support, not replace, clinician judgment: a clinician with a specific concern not captured by the five criteria (a worrisome mechanism, a patient who “doesn’t look right”) is not obligated to withhold imaging on a technicality just because the checklist came back clear. Embedding NEXUS in an EHR order set or clinical decision support tool is good practice for consistency, but the interface should still require the clinician to affirmatively document each of the five criteria rather than presenting a single pass/fail toggle — the latter invites exactly the kind of unreliable, un-auditable use the rule was built to prevent.

FAQ

What are the 5 NEXUS criteria?

No posterior midline cervical tenderness, no focal neurologic deficit, a normal level of alertness, no evidence of intoxication, and no painful distracting injury. All five must be present for a patient to be classified low-risk for cervical-spine injury under the rule.

Is the NEXUS criteria the same as the Canadian C-Spine Rule?

No. Both are real, validated cervical-spine clearance instruments used for the same general purpose, but they are structured differently (a flat checklist vs. a tiered, branching rule) and were validated in somewhat different populations. A 2003 head-to-head comparison in alert, stable trauma patients found the Canadian C-Spine Rule more sensitive and more specific than NEXUS in that specific population — see the comparison section above for the actual numbers and what they do and don’t imply.

How accurate is NEXUS for ruling out cervical spine injury?

In the original 2000 derivation and validation study of more than 34,000 patients, NEXUS performed with approximately 99% sensitivity, meaning it missed very few clinically significant injuries, with low specificity (meaning many patients flagged by the rule did not turn out to have an injury) — the expected trade-off for a rule-out tool. In the 2003 head-to-head comparison against the Canadian C-Spine Rule in a narrower population of alert, stable patients, NEXUS’s measured sensitivity and specificity were both lower than the CCR’s.

Can NEXUS be used to avoid imaging in every trauma patient?

No. It applies only to patients who meet all five low-risk criteria; any patient who fails even one criterion falls outside the rule’s low-risk group, and standard imaging workup should proceed. It is also not intended for penetrating trauma, and it cannot substitute for clinician judgment when a specific concern isn’t captured by the checklist.