Clinical decision rules for X‑ray: Ottawa, Canadian C-spine and NEXUS

Name: Ottawa ankle rules, Canadian C-spine rule and NEXUS explained
Creator: X-Ray Atlas editorial team
Published: 2026-10-25

Written by the X-Ray Atlas editorial team from the sources listed below · Medical review pending · Updated 25 October 2026 · Level: Beginner

A clinical decision rule is a short list of history and examination findings, tested in large prospective studies, that tells you whether an injured patient needs imaging. The best known are the Ottawa ankle and foot rules, the Ottawa knee rule, the Pittsburgh knee rule, the Canadian C-spine rule and the NEXUS criteria. They answer one question only: who can safely skip imaging? They do not tell you what the film shows, they apply only to the patients they were studied in, and in the UK they sit inside NICE guidance and your local policy. This page sets out each rule, the evidence behind it, and the patients it was never meant for.

1. What a decision rule is (and is not)

These rules were built to be highly sensitive: in the original studies they picked up almost every clinically important fracture, at the cost of only modest specificity. That trade-off is the whole point. A negative rule makes an important fracture very unlikely, so the patient can usually be managed without a film. A positive rule does not mean there is a fracture. It means imaging is indicated, and most patients imaged under a positive rule will have a normal film.

Three consequences follow:

The rule is only as good as the examination. The criteria depend on accurate palpation of specific bony points and an honest test of weight bearing or neck rotation. Sloppy examination breaks the rule.
The rule only applies to the study population. Each was tested in a defined group (for example alert, stable adults) and has stated exclusions. Outside that group you are back to clinical judgement and local guidance.
The rule never overrides concern. If you are worried despite a negative rule, discuss the patient with a senior colleague and follow local policy.

Pearl

Sensitivity and specificity in these papers refer to clinically important injuries, each study using its own definition set in advance. That is a deliberate design choice, but it is why "rule negative" means "an important fracture is very unlikely", not "no bony injury of any kind".

2. Ottawa ankle and foot rules

Stiell and colleagues derived the Ottawa ankle rules in 1992 from 750 adults with acute blunt ankle injury in two Canadian university emergency departments, then refined and prospectively validated them in 1993. The refined rule, as reproduced by later reviews and calculators, is:

Ottawa ankle and foot rules

Ankle X-ray series if there is pain in the malleolar zone and any of:

bone tenderness at the posterior edge or tip of the lateral malleolus (along its distal 6 cm);
bone tenderness at the posterior edge or tip of the medial malleolus (along its distal 6 cm);
inability to bear weight for four steps both immediately after the injury and in the emergency department.

Foot X-ray series if there is pain in the midfoot zone and any of:

bone tenderness at the base of the fifth metatarsal;
bone tenderness at the navicular;
inability to bear weight for four steps, as above.

The 1992 derivation version also used age 55 or over and included cuboid tenderness for the foot. These do not appear in the refined rule that is used today, so check which version a source is quoting.

Evidence. In the 1993 validation the refined rules were 100% sensitive for malleolar and midfoot fractures, with an estimated reduction of 34% in ankle series and 30% in foot series. Bachmann's 2003 systematic review (27 studies, 15,581 patients) found a pooled sensitivity of 97.6% (95% CI 96.4 to 98.9%), with a median specificity of about 31.5% and wide variation between studies. Applied to a fracture prevalence of 15%, a negative rule left a probability of fracture below 1.4%.

Who it applies to. The rules were derived in adults. Bachmann's review included children (pooled sensitivity 99.3%) but noted the evidence was sparse. NICE NG38 recommends the Ottawa ankle and foot rules to decide on X-ray in people over 5 years with a suspected ankle fracture. Radiopaedia advises caution in children under 6, and the child must be walking normally for the weight-bearing step to mean anything.

When the film is taken, see ankle fractures and foot fractures for how to read it.

3. Ottawa and Pittsburgh knee rules

The Ottawa knee rule was derived in 1995 in 1,047 adults with acute knee injury and validated in 1996 in 1,096 adults. A knee X-ray is indicated if any of the following is present:

Ottawa knee rule

age 55 or over;
tenderness at the head of the fibula;
isolated tenderness of the patella (no other bony tenderness);
inability to flex the knee to 90°;
inability to bear weight for four steps both immediately and in the emergency department.

In derivation the rule found all 68 fractures (sensitivity 1.0, 95% CI 0.95 to 1.0; specificity 0.54) and would have cut knee radiography by 28%. In validation it found all 63 clinically important fractures. A 2004 systematic review (6 studies, 4,249 adults) gave a pooled sensitivity of 98.5% and specificity of 48.6%. NICE NG38 recommends the Ottawa knee rules in people over 2 years with a suspected knee fracture.

The Pittsburgh knee rule was developed by Seaberg and Jackson in 1994. In the published rule, imaging is indicated after blunt trauma or a fall when the patient is either younger than 12 or older than 50, or unable to walk (weight bear) in the emergency department. In a 1998 multicentre comparison of 745 patients, it had a sensitivity of 99% and specificity of 60%, against 97% and 27% for the Ottawa rule in the same study.

Pitfall: specificity varies [VARIES]

Reported specificity of the Ottawa knee rule ranges from 54% in its derivation study and 48.6% in the pooled review down to 27% in the 1998 comparison. Different populations, examiners and fracture definitions explain much of this. Quote a range, not a single figure, and follow whichever rule your department uses.

For reading the film, including the fat–fluid level that betrays an intra-articular fracture, see knee fractures.

4. Canadian C-spine rule

The Canadian C-spine rule was derived by Stiell and colleagues (JAMA 2001) in 8,924 adults with blunt trauma to the head or neck, stable vital signs and a Glasgow Coma Scale (GCS) score of 15. 151 (1.7%) had a clinically important C-spine injury. The rule asks three questions in order:

Canadian C-spine rule

Any high-risk factor that mandates imaging? Age 65 or over; a dangerous mechanism; or paraesthesia in the limbs. Dangerous mechanism, as defined in the authors' later validation, means a fall from more than 3 ft (91 cm) or five stairs, an axial load to the head (for example diving), a high-speed motor vehicle collision (over 100 km/h), a rollover or ejection, a motorised recreational vehicle, or a bicycle collision. If yes: image.
Any low-risk factor that allows safe assessment of range of movement? A simple rear-end collision; sitting in the emergency department; walking at any time since the injury; delayed onset of neck pain; or no midline C-spine tenderness. If none: image. A rear-end collision is not "simple" if the car was pushed into oncoming traffic, hit by a bus, large truck or high-speed vehicle, or rolled over.
Can the patient actively rotate the neck 45° to the left and to the right? If not: image. If yes: no imaging is needed.

In derivation the rule had 100% sensitivity (95% CI 98 to 100%) and 42.5% specificity.

Who it does not apply to. In Stiell's validation work the rule is used only in alert (GCS 15), stable adults with blunt trauma in the previous 48 hours. Patients were excluded if they were under 16, had penetrating trauma or acute paralysis, had known vertebral disease (ankylosing spondylitis, rheumatoid arthritis, spinal stenosis or previous cervical spine surgery), or were returning for reassessment or referred from another hospital. Patients with these features need clinical judgement and local trauma guidance, not the rule.

For how to read the film or CT in those who do need imaging, see cervical spine X-ray.

5. NEXUS, and how it compares

The NEXUS (National Emergency X-Radiography Utilization Study) low-risk criteria were validated by Hoffman and colleagues in 34,069 patients who had cervical spine radiography after blunt trauma at 21 US centres (NEJM 2000). A patient is low risk, and needs no imaging, only if all five are met:

NEXUS low-risk criteria

no posterior midline cervical tenderness;
no focal neurological deficit;
normal alertness;
no intoxication;
no painful, distracting injury.

NEXUS identified all but 8 of 818 injured patients (sensitivity 99.0%, 95% CI 98.0 to 99.6%), with a negative predictive value of 99.8% but a specificity of only 12.9%. Imaging could have been avoided in 12.6% of patients. NEXUS enrolled children: a separate analysis of the 3,065 patients under 18 found no missed injuries, but there were only 88 children under 2, and the authors advised caution in infants and toddlers.

Head to head. In 2003 Stiell's group compared the two rules in 8,283 alert, stable patients in nine Canadian emergency departments. Excluding the 10.2% of patients in whom physicians did not test neck rotation, the Canadian rule was more sensitive (99.4% vs 90.7%) and more specific (45.1% vs 36.8%), and would have led to fewer radiographs (55.9% vs 66.6%). The Canadian rule would have missed 1 important injury and NEXUS 16. Both the derivation and the comparison were run by the team that developed the Canadian rule, in Canadian emergency departments, which is worth remembering when you read the result.

	Canadian C-spine rule	NEXUS
Population studied	Alert (GCS 15), stable adults with blunt head or neck trauma	Blunt trauma patients of any age who had C-spine imaging
Uses age	Yes (65 or over is high risk)	No
Uses mechanism	Yes	No
Tests neck movement	Yes (active rotation 45° each way)	No
Accuracy in the 2003 comparison	Sensitivity 99.4%, specificity 45.1%	Sensitivity 90.7%, specificity 36.8%

6. How UK guidance uses the rules

Cervical spine, adults. NICE NG41 (spinal injury) tells clinicians to assess C-spine risk using the Canadian C-spine rule, and to do CT in adults (16 or over) when the rule indicates imaging. NICE NG232 (head injury) uses the same high-risk factors, with a fall threshold of more than 1 m or 5 stairs, and adds others such as GCS 12 or less and intubation. For adults the imaging test is CT, not plain films.
Cervical spine, children. NICE NG41 notes that applying the Canadian rule to children is difficult and that the child's developmental stage must be considered. For under-16s with head injury, NG232 sets out when to do CT and when 3-view cervical spine X-rays are appropriate. Paediatric C-spine decisions should follow these recommendations and local paediatric policy.
Ankle, foot and knee. NICE NG38 (non-complex fractures) recommends the Ottawa ankle and foot rules in people over 5 and the Ottawa knee rules in people over 2.

NEXUS is widely used in North America but is not the rule NICE recommends for adults. Always check the current version of the guideline and your department's policy, which may add local criteria.

7. Pitfalls

Palpating the wrong place. The Ottawa ankle rule depends on the posterior edge and tip of each malleolus. Tenderness only over the anterior ligaments is not a positive rule.
Testing neck rotation when it isn't safe. In the Canadian rule, rotation is tested only when there is no high-risk factor and at least one low-risk factor. NICE NG232 makes the same point.
Using a rule outside its population: intoxicated, confused or uncooperative patients, those under 16 (for the Canadian C-spine rule), and patients with ankylosing spondylitis or previous neck surgery.
Treating "rule positive" as "fracture". Most rule-positive films are normal. The film still needs a systematic read: see how to read a skeletal X-ray.
Forgetting the safety net. A patient managed without imaging should get clear advice to return if symptoms do not settle, as set out in local policy.

Red flags: act now

Do not rely on a decision rule; escalate

Any new neurological deficit, limb weakness or paraesthesia after neck injury: keep the spine protected and follow trauma and spinal injury guidance.
Reduced GCS, intoxication or a distracting injury that makes the examination unreliable.
Known axial spondyloarthritis (ankylosing spondylitis) or other predisposing spinal disease: these patients fall outside the Canadian rule, and NICE NG232 lists such conditions as a reason for CT in adults with a head injury and neck pain.
An obvious deformity or a limb that looks threatened (cold, pale or numb): this needs urgent senior assessment and imaging whatever the rule says.
An open wound over a possible fracture.

What to focus on, by role

Role	Priorities
Medical student	Learn each rule exactly, including the order of the Canadian C-spine questions. Be able to state who each rule excludes.
Nurse / ENP	Examine the exact bony points, test four steps honestly, and know when a rule cannot be used. Follow your department's protocol for nurse-requested imaging.
Junior doctor	Apply the Canadian C-spine rule as NICE NG41 and NG232 set out, request CT rather than films for adults who need C-spine imaging, and document the rule and its result.
Radiographer	Understand why a referral meets the rule, and query requests that fall outside it in line with IR(ME)R justification and local policy.
Advanced / FRCR	Know the derivation and validation evidence, the 2003 head-to-head comparison and its limits, and how the rules are built into NICE guidance.

Related: how to describe a fracture, the glossary, and practice cases in lower limb and spine.

References

Stiell IG, Wells GA, Vandemheen KL, et al. The Canadian C-spine rule for radiography in alert and stable trauma patients. JAMA 2001;286(15):1841–8. doi:10.1001/jama.286.15.1841
Stiell IG, Clement CM, O'Connor A, et al. Multicentre prospective validation of use of the Canadian C-Spine Rule by triage nurses in the emergency department. CMAJ 2010;182(11):1173–9 (rule definitions and exclusions). pmc.ncbi.nlm.nih.gov
Hoffman JR, Mower WR, Wolfson AB, et al. Validity of a set of clinical criteria to rule out injury to the cervical spine in patients with blunt trauma. N Engl J Med 2000;343(2):94–9. doi:10.1056/NEJM200007133430203. Paediatric subgroup: Viccellio P, et al. Pediatrics 2001;108(2):e20. doi:10.1542/peds.108.2.e20
Stiell IG, Clement CM, McKnight RD, et al. The Canadian C-spine rule versus the NEXUS low-risk criteria in patients with trauma. N Engl J Med 2003;349(26):2510–8. doi:10.1056/NEJMoa031375
Stiell IG, Greenberg GH, McKnight RD, et al. A study to develop clinical decision rules for the use of radiography in acute ankle injuries. Ann Emerg Med 1992;21(4):384–90; and Decision rules for the use of radiography in acute ankle injuries: refinement and prospective validation. JAMA 1993;269(9):1127–32. doi:10.1001/jama.269.9.1127. Rule wording also checked against Radiopaedia: Ottawa ankle rules.
Bachmann LM, Kolb E, Koller MT, et al. Accuracy of Ottawa ankle rules to exclude fractures of the ankle and mid-foot: systematic review. BMJ 2003;326:417. pmc.ncbi.nlm.nih.gov
Stiell IG, et al. Derivation of a decision rule for the use of radiography in acute knee injuries. Ann Emerg Med 1995;26(4):405–13; Prospective validation. JAMA 1996;275(8):611–5; Bachmann LM, et al. Ann Intern Med 2004;140(2):121–4; Seaberg DC, Jackson R. Am J Emerg Med 1994;12:541–3; Seaberg DC, et al. Multicenter comparison of two clinical decision rules for the use of radiography in acute, high-risk knee injuries. Ann Emerg Med 1998;32(1):8–13. doi:10.1016/s0196-0644(98)70092-7
NICE. Spinal injury: assessment and initial management (NG41); Head injury: assessment and early management (NG232); Fractures (non-complex): assessment and management (NG38). NG41, NG232, NG38

For education only. Not a substitute for a formal report, senior review or your local policy. Never use this page to make decisions about a real patient.