NEXUS or Canadian rule: cervical spine imaging
Table of contents
A 38-year-old man arrives at the emergency department after a low-speed rear-end collision. He arrived walking, is wearing a cervical collar placed by the rescue team and complains of mild neck pain. The neurological exam is normal. The resident asks the attending physician whether she needs to order a CT scan of the cervical spine or whether she can remove the collar.
This is one of the most common questions in the emergency room, and two decision rules were created to answer it: the NEXUS criteria, born from a large North American study, and the Canadian cervical spine rule. Both try to separate those who can skip imaging from those who need it. This text compares the two rules, shows what studies from several countries say about their performance and discusses where older adults and children fall outside the logic of each one.
Key points
- NEXUS and the Canadian rule serve to identify who can skip cervical spine imaging after trauma.
- Both are very sensitive, and the specificity of both is low.
- The Canadian rule calls for imaging in all older adults, and this point is under debate.
- In older adults after a fall and in young children, the rules need extra care.
- A rule only changes practice when it enters the service workflow.
Why we need a rule
Cervical spine injuries are rare among trauma patients who arrive awake and stable, but an injury that goes unnoticed can cause paralysis. This imbalance leads many services to order imaging for nearly everyone. The result is a large volume of normal exams, with radiation exposure, longer length of stay and occupation of the CT scanner.
Baker and colleagues described an emergency department that performed more than 40 000 CT scans per year, with use above 80 out of every 100 of the scanner's capacity (Baker and colleagues, 2020). In this setting, each unnecessary exam delays another that would be needed. The text on overdiagnosis discusses the harms of investigating too much, and the text on clinical decision-making shows how to choose exams based on what changes management.
What a decision rule does
A clinical decision rule combines findings from the history and the exam into a sequence that leads to a recommendation. To be useful, it needs high sensitivity, because the most serious error is to skip imaging in someone who has an injury. Specificity matters less, but it cannot be so low that the rule stops reducing exams.
The rules for the cervical spine were derived in specific populations, and this defines the limits of use. They apply to patients who are awake, stable and have blunt trauma, and each one has its own exclusion criteria. Applying the rule outside this population is one of the most common failures in practice.
The collar also has effects
The cervical collar protects the spine while the injury has not been ruled out, but it is not harmless. It is uncomfortable, makes airway assessment harder, can cause skin injuries in those immobilized for many hours and increases the patient's anxiety. Therefore, the decision about imaging is also a decision about how long the patient will stay in the collar. A rule that safely clears those at low risk gives comfort back to the patient and frees the team for the problems that actually threaten life.
That is why the decision rule has a dual role. It avoids unnecessary exams and, at the same time, allows earlier removal of the collar in those who do not need it. The man in the opening scene benefits from a quick and safe decision, whether it is imaging or clinical clearance.
How the two rules work
NEXUS
The NEXUS criteria skip imaging when the patient meets five conditions at the same time: no pain on palpation of the posterior midline of the neck, no focal neurological deficit, normal level of alertness, no intoxication and no distracting painful injury. Baratloo and colleagues summarize these items as midline tenderness, intoxication, alertness, focal neurological deficit and distracting injury (Baratloo and colleagues, 2023). If any of these items fails, the recommendation is to order imaging.
The advantage of NEXUS is its simplicity, with only five criteria to check before removing the collar. They are questions that are easy to remember and to apply at the bedside. The difficulty lies in terms such as "distracting injury" and "normal level of alertness", which depend on the interpretation of the examiner and vary greatly between professionals.
NEXUS in the medical record
Baker and colleagues put a form based on NEXUS into the electronic medical record and trained the professionals in the criteria (Baker and colleagues, 2020). The proportion of potentially inappropriate cervical spine CT scans fell from 19 in 42 to 90 in 403 (Baker and colleagues, 2020). After the intervention, the difference between physicians and other professionals who ordered imaging disappeared (Baker and colleagues, 2020).
This result shows that the rule gains strength when it appears at the moment of the decision. A rule kept in memory competes with the fatigue and the rush of the shift, and the same rule built into the imaging order becomes part of the workflow.
The Canadian rule
The Canadian rule follows three questions in sequence. The first looks for high-risk factors, such as age 65 or older, a dangerous mechanism or limb paresthesias, and any of them indicates imaging (Baratloo and colleagues, 2023). The second looks for low-risk factors that allow mobility to be assessed safely, such as a simple rear-end collision, a patient sitting in the emergency department or walking since the accident. The third tests whether the patient can turn the neck 45 degrees to each side. Baratloo and colleagues describe the rule as the sum of age, trauma mechanism, paresthesias, the patient's position and ambulation, and the range of neck rotation (Baratloo and colleagues, 2023).
This structure is longer than NEXUS, but it gives weight to mechanism and age. The man in the opening scene, who arrived walking after a simple rear-end collision, has a low-risk factor and can have his mobility tested, as long as there are no high-risk factors.
The rule in the real world
Stiell and colleagues ran a cluster randomized trial in 12 hospitals in Canada, with 11,824 alert and stable adults with blunt head or neck trauma (Stiell and colleagues, 2009). In the hospitals that received the active implementation strategy, with education, institutional policy and reminders in the imaging order, the cervical spine imaging rate went from 61.7 to 53.3 per 100 (Stiell and colleagues, 2009). In the control hospitals, the rate rose from 52.8 to 58.9 per 100 (Stiell and colleagues, 2009). The authors reported that the reduction occurred without missed injuries or harm to patients (Stiell and colleagues, 2009).
What the comparisons say
Several studies applied both rules to the same group of patients. Vazirizadeh-Mahabadi and colleagues did a systematic review with meta-analysis and included 5 studies with direct comparison (Vazirizadeh-Mahabadi and colleagues, 2023). The pooled sensitivity was 0.899 for NEXUS and 0.987 for the Canadian rule (Vazirizadeh-Mahabadi and colleagues, 2023). The specificity was 0.398 for NEXUS and 0.167 for the Canadian rule (Vazirizadeh-Mahabadi and colleagues, 2023). The authors concluded that the Canadian rule appeared to have better screening accuracy, and that the low specificity of both means many people continue to receive unnecessary imaging (Vazirizadeh-Mahabadi and colleagues, 2023).
The larger meta-analysis
Baratloo and colleagues gathered 35 articles, with about 70,000 patients assessed by NEXUS and 33,000 by the Canadian rule (Baratloo and colleagues, 2023). For any cervical injury, sensitivity was 0.94 for NEXUS and 1.00 for the Canadian rule (Baratloo and colleagues, 2023). For clinically important injury, the area under the curve was 0.78 for NEXUS and 0.94 for the Canadian rule (Baratloo and colleagues, 2023). The authors concluded that both tools serve to rule out injury in low-risk patients, with better performance from the Canadian rule, and they noted that the specificity of both is low (Baratloo and colleagues, 2023).
Even with low specificity, applying these rules can still greatly reduce the number of radiographs taken in emergency departments, according to the same review (Baratloo and colleagues, 2023). The result matches that of the earlier meta-analysis: the Canadian rule flags almost every patient with injury, at the cost of also flagging many without injury.
The differences between studies have known explanations. Populations with different profiles, different definitions of important injury, small samples and variation in how the criteria are applied change the numbers. The text on pretest probability helps explain why the same test yields different results in different places.
How to read sensitivity and specificity here
In a rule for forgoing tests, high sensitivity is the safety guarantee. It says that almost all patients with injury will be flagged for imaging. Low specificity says something else: many patients without injury will also be flagged, and the rule reduces tests less than one would like.
The combination of the two shows why no rule eliminates the normal CT scan. It only removes from the queue a portion of very low-risk patients. The real gain depends on how many low-risk patients the department sees and on how much the team trusts the rule to not order the test anyway. The text on base rate in diagnosis explains why the rarity of the injury weighs so heavily in this calculation.
| Study | Population | Main finding | Source |
|---|---|---|---|
| Implementation trial | 11,824 adults in 12 hospitals | Imaging fell from 61.7 to 53.3 per 100 | (Stiell and colleagues, 2009) |
| Meta-analysis | 5 studies with direct comparison | Sensitivity of 0.899 for NEXUS and 0.987 for the Canadian rule | (Vazirizadeh-Mahabadi and colleagues, 2023) |
| Broad meta-analysis | 35 articles | Area under the curve of 0.78 for NEXUS and 0.94 for the Canadian rule for important injury | (Baratloo and colleagues, 2023) |
| Modified NEXUS | 795 older adults after a fall | 11 cervical injuries, all identified | (Tran and colleagues, 2016) |
| Japanese rule | 927 patients with follow-up | Sensitivity of 92.1 per 100 for the initial rule | (Inagaki and colleagues, 2018) |
Other rules
Inagaki and colleagues derived a rule for cervical CT in two emergency departments in Japan, from August 2012 to March 2014, with patients aged 16 or older with head or neck trauma (Inagaki and colleagues, 2018). Of the 1192 patients included, 927 completed follow-up and 38 had cervical injury (Inagaki and colleagues, 2018). The initial rule had sensitivity of 92.1 per 100 and specificity of 58.6 per 100, and the authors refined the rule with the Glasgow scale, neck pain, neurological deficit and mechanism (Inagaki and colleagues, 2018). The new version still needs independent validation (Inagaki and colleagues, 2018).
Older adults: age as a criterion
The Canadian rule calls for imaging for everyone aged 65 or older. Rodriguez and Greenberg note that this is the most used clearance rule and that other evidence-based recommendations for the cervical spine in older adults are lacking (Rodriguez and Greenberg, 2025). They commented on a study that reviewed 9455 CT scans in 7114 patients aged 65 or older, from 2018 to 2023, in two hospitals (Rodriguez and Greenberg, 2025). There were 192 scans with cervical fracture, and 28 of these were classified as asymptomatic (Rodriguez and Greenberg, 2025).
Commentators pointed out important limitations of this study, such as the uncertain denominator, the lack of details on the indications for the exams and the difficulty of defining symptoms in a chart review (Rodriguez and Greenberg, 2025). They concluded that the data are still insufficient to change practice, given the harm of a missed cervical injury (Rodriguez and Greenberg, 2025).
High injuries and trivial falls
The injury pattern also changes with age. Wang and colleagues reviewed a trauma registry with 12,805 patients, from 2006 to 2010, and found 726 with cervical fracture (Wang and colleagues, 2013). Of the 139 older adults with fracture, 75 had a C2 fracture, against 129 of the 587 younger patients (Wang and colleagues, 2013). The authors concluded that older adults tend to have high cervical fractures, regardless of mechanism, and that a fall from standing height can cause significant injuries (Wang and colleagues, 2013).
The modified NEXUS for older adults
Tran and colleagues tested a modified NEXUS in older adults who fell and who were seen outside the trauma room of a level 1 center (Tran and colleagues, 2016). In the modified version, only signs of trauma to the head or neck counted as a distracting injury, and the usual mental state counted as normal alertness (Tran and colleagues, 2016). Of the 795 patients analyzed, with a mean age of 83.6 years, 11 had a cervical injury, and the modified rule identified all of them (Tran and colleagues, 2016).
This adaptation shows the logic of the reasoning. The older adult with dementia may always have partial orientation, and calling that an altered level of alertness leads to imaging in almost all of them. Defining normal by the patient's previous state makes the rule applicable to them.
When the CT scan is normal
Cushing and colleagues reviewed 87 patients aged 65 or older who underwent cervical spine MRI after a normal CT scan following a fall from standing height (Cushing and colleagues, 2021). Of these, 10 had traumatic changes on MRI, and one patient needed surgery for an unstable injury (Cushing and colleagues, 2021). The authors call for better criteria to define when MRI should be performed in this group (Cushing and colleagues, 2021).
In the practice of many services, a good-quality negative CT scan has become the step that allows the collar to be removed in most patients. Motor symptoms or disagreement between the clinical picture and the imaging call for evaluation by a spine surgeon, and that conversation is worth more than repeating exams out of insecurity.
Children: another anatomy
A child's cervical spine is different from an adult's. The head is proportionally larger and the spine is more flexible, which changes the type of injury. Gopinathan and colleagues reviewed 67 articles on pediatric cervical injury and reported that children have atlantoaxial injuries 2.5 times more often than adults (Gopinathan and colleagues, 2018). Clinical clearance criteria can, in general, be applied to children over 2 years old, with caution in younger ones (Gopinathan and colleagues, 2018).
Phillips and colleagues described the protocol of a prospective study in 14 emergency departments in Australia, New Zealand and Singapore, with children under 16 years old with possible cervical injury after blunt trauma (Phillips and colleagues, 2025). The study will compare the performance of NEXUS, the Canadian rule and a new pediatric rule from the PECARN network (Phillips and colleagues, 2025). While these results arrive, the choice of rule in children remains more uncertain than in adults.
What changes in pediatric reasoning
Young children cannot always report pain or cooperate with the mobility exam, and this makes the criteria less reliable. Parental participation, observation of behavior and reassessment gain weight. The text on interprofessional clinical reasoning shows how nursing, physicians and family can add observations in this type of assessment.
How to choose in practice
Both rules are safe when applied to the right population, and the choice depends on the service. The Canadian rule tends to be more sensitive and reduces more tests when well implemented, as the Canadian trial showed (Stiell and colleagues, 2009). NEXUS is simpler and can be easily incorporated into the medical record, as in the study by Baker (Baker and colleagues, 2020).
- Confirm that the patient is awake, stable and had blunt trauma.
- Choose a single rule for the service and train the whole team in it.
- Embed the rule in the imaging request or in the medical record.
- In older adults, consider the usual mental state and the chance of high fractures.
- In small children, add the parents' observation and reassessment.
The text on diagnostic uncertainty helps decide what to do when the rule does not apply, and the text on how to organize reasoning in the medical record shows how to document the decision to skip imaging.
Where reasoning often fails
Three failures appear often. The first is applying the rule to those outside it, such as the intoxicated patient or one with a distracting injury. The second is calling "mild pain" a midline pain that should indicate imaging. The third is treating a normal CT scan in an older adult with a neurological deficit as the end of the investigation.
Discussing these cases in morbidity and mortality sessions helps the team see where the rule was misapplied. The text on learning based on clinical cases shows how to train the application of rules with varied scenarios.
The reasoning behind the rule
Applying a decision rule seems like an automatic gesture, but each item calls for judgment. Deciding whether an arm fracture distracts the patient, whether slurred speech comes from alcohol or from head trauma, or whether the neck pain is in the midline or to the side of it requires careful examination. The rule organizes the questions and the examiner remains responsible for the answers.
A good habit is to record in the medical record each criterion with the finding that supports it. Instead of noting only that the patient meets the rule, the team writes that the patient is alert, denies pain on midline palpation, turns the neck to both sides and has a normal neurological examination. This record helps whoever takes over the shift, makes auditing easier and makes the reasoning visible to students.
It is also worth remembering that the rule answers a specific question: who can skip imaging. It leaves aside other important questions, such as the choice between radiography, CT and MRI, the need for immobilization during transport and referral to the specialist. Mixing these questions usually leads to repeated tests and collars kept on longer than necessary.
Finally, the rule works best when the whole team knows it. The technician who positions the patient, the nurse who measures pain and the doctor who decides speak the same language, and the decision becomes faster and safer for everyone.
How NEXO helps train this reasoning
In NEXO, the case of the day is fictional, with a real clinical presentation and up to six progressive clues, and you propose the diagnosis by searching for the ICD-10 code.
The limits of use of the game are described on the NEXO economy page. If you research the teaching of clinical reasoning, the page for researchers explains how to get in touch. Professionals who write or review cases find information in curation and in creators, and institutions can learn about the proposal on the partnerships page. Training with differential diagnosis and clues and searching for the right ICD-10 code complete this training.
Frequently asked questions
Which is more sensitive, NEXUS or the Canadian rule?
In the meta-analysis of 5 studies, the combined sensitivity was 0.987 for the Canadian rule and 0.899 for NEXUS (Vazirizadeh-Mahabadi and colleagues, 2023).
Does the Canadian rule reduce tests in practice?
It does when well implemented, and in the Canadian trial the imaging rate fell from 61.7 to 53.3 per 100 in hospitals with an active strategy (Stiell and colleagues, 2009).
Does every older adult with trauma need cervical imaging?
The Canadian rule recommends imaging for those aged 65 or older, and commentators consider current data insufficient to change this practice (Rodriguez and Greenberg, 2025).
Do the rules apply to children?
The clinical criteria can generally be applied to children over 2 years old, with caution in younger ones (Gopinathan et al., 2018).
Does a normal CT end the investigation in older adults?
It does in most cases, but among 87 older adults who had an MRI after a normal CT, 10 had traumatic changes (Cushing et al., 2021).
Take-home messages
- NEXUS and the Canadian rule help safely skip cervical imaging.
- Both are very sensitive and not very specific.
- A rule only changes practice when it enters the service workflow.
- In older adults, usual mental status and high fractures call for attention.
- In small children, observation and reassessment gain weight.
References
- Baker M, Jaeger C, Hafley C, et al. Appropriate CT cervical spine utilisation in the emergency department. BMJ Open Qual. 2020. PMID 33028655.
- Baratloo A, Ahmadzadeh K, Forouzanfar MM, et al. NEXUS vs. Canadian C-Spine Rule (CCR) in Predicting Cervical Spine Injuries; a Systematic Review and Meta-analysis. Arch Acad Emerg Med. 2023. PMID 37840870.
- Cushing CH, Holmes JF, Tyler KR. Cervical Spine Injuries in Older Patients with Falls Found on Magnetic Resonance Imaging After Computed Tomography. West J Emerg Med. 2021. PMID 34546897.
- Gopinathan NR, Viswanathan VK, Crawford AH. Cervical Spine Evaluation in Pediatric Trauma: A Review and an Update of Current Concepts. Indian J Orthop. 2018. PMID 30237606.
- Inagaki T, Kimura A, Makishi G, et al. Development of a new clinical decision rule for cervical CT to detect cervical spine injury in patients with head or neck trauma. Emerg Med J. 2018. PMID 30032123.
- Phillips N, Askin GN, Davis GA, et al. Prospective observational study to assess the performance accuracy of clinical decision rules in children presenting to emergency departments with possible cervical spine injuries: the Study of Neck Injuries in Children (SONIC). BMJ Open. 2025. PMID 40316355.
- Rodriguez R, Greenberg S. Is older age an appropriate criterion alone for ordering cervical spine CT after trauma. Acad Emerg Med. 2025. PMID 39119986.
- Stiell IG, Clement CM, Grimshaw J, et al. Implementation of the Canadian C-Spine Rule: prospective 12 centre cluster randomised trial. BMJ. 2009. PMID 19875425.
- Tran J, Jeanmonod D, Agresti D, et al. Prospective Validation of Modified NEXUS Cervical Spine Injury Criteria in Low-risk Elderly Fall Patients. West J Emerg Med. 2016. PMID 27330655.
- Vazirizadeh-Mahabadi M, Yarahmadi M. Canadian C-spine Rule versus NEXUS in Screening of Clinically Important Traumatic Cervical Spine Injuries; a systematic review and meta-analysis. Arch Acad Emerg Med. 2023. PMID 36620739.
- Wang H, Coppola M, Robinson RD, et al. Geriatric Trauma Patients With Cervical Spine Fractures due to Ground Level Fall: Five Years Experience in a Level One Trauma Center. J Clin Med Res. 2013. PMID 23519239.