X-ray interpretation in the emergency department — the systematic approach

Plain radiography is the highest-volume imaging modality in the emergency department. It is fast, cheap, widely available, and low-radiation, and it is the first-line test for the fractured limb, the breathless chest, the swallowed foreign body and the suspected perforation. The Fellowship examiner is not testing whether a candidate can name a favourite film; the test is whether a candidate can apply a reproducible systematic method to any radiograph, in any patient, under pressure, and not miss the tube in the bronchus, the free gas under the diaphragm, or the paediatric growth-plate injury. Pattern recognition fails on the rotated portable film; a system does not. The interpreter — not the radiologist's later report — makes the disposition. [1]

When to use the framework and the principles of a systematic read

Every ED patient with a suspected fracture, a chest symptom or sign, an in-dwelling tube or line, or a suspected bowel obstruction or perforation is interpreted against a plain film. The radiograph is read by the treating clinician at the time of decision — the formal radiologist report is a quality check, not a handover. A scoping review of chest-radiograph interpretation confirms that a structured, checklist-driven read is more reproducible than an unstructured scan, and that the systematic method is teachable.[8] Five principles govern the read. First, assess the technical quality before any pathology — a film that is rotated, under-penetrated or poorly inspired cannot be cleared. Second, use the same ordered sequence every compartment, every time, so nothing is skipped under cognitive load. Third, interpret in parallel with resuscitation — a tension pneumothorax is decompressed, not described. Fourth, read with the clinical question and compare with prior films; a borderline finding that is new is far more significant than one that is chronic. Fifth, state the limitation — a single radiograph excludes almost nothing.

Technical quality — projection, penetration, rotation, inspiration

Before any pathology is sought, the film must pass a quality gate. Projection is named for the direction the beam travels relative to the detector — a PA chest is taken postero-anterior with the beam entering the back, the heart sits close to the detector and is minimally magnified, whereas a portable AP chest magnifies the heart and widens the mediastinum, so the cardiomediastinal contour is never declared abnormal on a single AP film alone. Penetration is adequate when the lower thoracic vertebral bodies are just visible through the cardiac shadow; an under-penetrated film hides the retrocardiac lung and the posterior costophrenic recesses. Rotation is judged from the medial ends of the clavicles, which should sit symmetrically about the spinous process; a rotated chest obliques the hilar contours and distorts the heart, and a rotated c-spine makes the facet joints asymmetric and falsely opens the disc spaces. Inspiration is adequate when five or six anterior ribs intersect the diaphragm, or nine or ten posterior ribs — a poor inspiratory effort crowds the lung bases, elevates the diaphragm, and widens the cardiomediastinal silhouette, simulating disease. [1]

The chest radiograph method — airway to tubes

The chest is read in a fixed sequence so that no compartment is missed. The mnemonic encodes the compartments, not the order of importance — the order is the discipline. [1]

The silhouette sign is the single most powerful localiser on a chest radiograph: an opacity that abuts and erases a structure is in contact with it (anterior to the heart border is lingula or middle lobe; the diaphragm is a lower lobe), whereas an opacity that preserves the border lies away from it. Collapse (atelectasis) shows volume loss — shift of the mediastinum or fissure toward the lesion, with crowded vessels — whereas consolidation preserves volume and shows an air bronchogram. [1]

The musculoskeletal method — ABCCS

The limb and joint radiograph is read with a deliberate, layered method that starts with film adequacy and ends with a stated summary. The two-view minimum is non-negotiable — a single anterior-posterior film misses a posterior shoulder dislocation, an undisplaced scaphoid fracture and a subtle radial head fracture. [1]

Salter-Harris — the paediatric physeal classification

The growing long bone has a radiolucent cartilage growth plate (the physis) between the epiphysis (the secondary ossification centre at the joint end) and the metaphysis (the flared end of the shaft). A fracture that crosses the physis can arrest growth, and the Salter-Harris grade predicts that risk — higher grades carry a higher rate of premature physeal closure and growth disturbance.[7]

The paediatric skeleton also produces patterns that have no adult equivalent. A buckle (torus) fracture is a compression failure — the cortex buckles and impinges but does not break through; the distal radius after a fall onto the outstretched hand is the archetype, managed in a removable splint. A greenstick fracture breaks one cortex and leaves the opposite intact (incomplete, plastic paediatric bone). A plastic (bowing) deformation bows a long bone without a discrete fracture line. An avulsion fractures a fragment at a tendon or ligament insertion — the tibial spine (anterior cruciate ligament), the patellar sleeve (quadriceps), the medial epicondyle of the elbow — and is the sports-injury archetype in adolescents. [1]

Fracture descriptors — the language of the read

The summary step requires the fracture to be described in the standard language. A transverse fracture is perpendicular to the long axis (a direct blow, stable); an oblique runs at an angle (a mix of compression and bending); a spiral winds around the shaft (a torsion mechanism — in a non-ambulant child, a spiral long-bone fracture is a non-accidental injury red flag). A comminuted fracture produces more than two fragments; a segmental fracture produces a separate shaft fragment between two fracture lines (a high-energy open-injury pattern); a butterfly fragment is a triangular wedge torn off by bending. A fracture is open (compound) when the wound communicates with the outside — antibiotics and tetanus cover are mandatory. A pathologic fracture traverses abnormal bone (metastasis, myeloma, a simple bone cyst). [1]

Differential diagnosis — normal, abnormal, artefact

A radiograph is a two-dimensional shadow of a three-dimensional patient, and the eye is fooled. Every film is judged against four possibilities. [1]

The decision rules — when to image in the first place

The decision to image is itself an examinable skill, and the high-yield rules were derived and validated to reduce unnecessary radiography without missing injury. [1]

The Ottawa ankle rule has a sensitivity approaching 100% and a near-perfect negative predictive value for clinically significant malleolar fracture, reducing ankle radiography by around a third without a missed injury.[1] The Ottawa knee rule shares the same near-perfect sensitivity for clinically significant knee fracture.[2] The Pittsburgh rule performs comparably and is a recognised alternative.[3] For the cervical spine, NEXUS requires all five low-risk criteria to be met before imaging is withheld (sensitivity 99.6% for clinically important injury)[4] and the Canadian C-spine rule, which is more sensitive and more specific, applies to alert and stable trauma patients.[5]

Image-to-action management thresholds

The interpretation does not stop at a label — several findings immediately trigger a drug, a procedure or a referral. The doses are committed to memory because the unstable patient cannot wait for a lookup. [1]

Tubes, lines and the portable trauma film

The intubated trauma patient is radiographed portably, supine and AP, which magnifies the heart and widens the mediastinum — a widened mediastinum on an AP trauma film is a prompt to a CT aortogram, not a diagnosis of dissection. The endotracheal tube tip should sit 2 to 3 cm above the carina (around the T4–T5 level); too high risks extubation with neck movement, too low risks main-stem intubation (the right main bronchus is steeper, shorter and more vertical, so a tube advanced too far goes right). The central venous catheter tip should lie in the cavoatrial junction, below the inferior border of the right main bronchus and above the right atrium, and not curl in the internal jugular or the contralateral subclavian. The nasogastric tube must not coil in the oropharynx and the tip must lie below the diaphragm in the stomach — a tube in a main bronchus is a never-event, withdrawn before any use. [1]

Special populations

The paediatric skeleton has open physes (Salter-Harris applies), accessory ossicles that mimic fracture, and plastic/bowing and greenstick patterns; the non-accidental-injury patterns are examinable — a posterior rib fracture near the costovertebral junction, a corner (bucket-handle) fracture of a long-bone metaphysis, a spiral long-bone fracture in a non-ambulant child, and multiple fractures of differing ages. The elderly patient has osteopenia, so a low-energy mechanism (a fall from standing) produces a hip, a distal radius, a proximal humerus or a vertebral compression fracture, and a normal radiograph in osteopenic bone does not exclude an occult fracture — an early MRI is the test. The pregnant patient is imaged under the ALARA principle; a single chest radiograph delivers around 0.01 millisieverts, far below the deterministic threshold for fetal harm at 100 millisieverts, and is justified when clinically indicated, with the abdomen shielded. The intubated and ventilated patient is radiographed portably AP and supine, with a magnified mediastinum and pooled posterior fluid — free intraperitoneal gas is sought on a supine film by the Rigler sign (gas on both sides of the bowel wall) or the football sign. [1]

Common errors and pitfalls

The recurring errors are reading only one view (the missed scaphoid, the missed posterior shoulder dislocation); the satisfaction of search — a second abnormality is missed once the first is found; declaring a widened mediastinum abnormal on a single rotated AP trauma film; clearing a cervical spine on an inadequate film (the C7–T1 junction must be visible); missing the nasogastric tube in the bronchus; mistaking a skin fold for a pneumothorax (the deep sulcus sign on a supine film is a true tension pneumothorax until proven otherwise); accepting a radiologist report in lieu of the clinician's own read; and forgetting that a negative radiograph excludes almost nothing — the early scaphoid fracture is occult for up to two weeks, early osteomyelitis is occult, and a pulmonary embolism produces a normal chest radiograph more often than an abnormal one. [1]

Evidence and regional guidelines

The imaging decision rules are among the most validated instruments in emergency medicine: the Ottawa ankle and knee rules have near-perfect sensitivity and have safely reduced radiography worldwide,[1][2] with the Pittsburgh knee rule as a comparable alternative.[3] The NEXUS and Canadian C-spine rules together define the cervical-spine imaging gate.[4][5] For paediatric elbow injury, the radiographic fat-pad sign is a high-yield occult-fracture marker — a meta-analysis of children with a positive posterior fat pad and no visible fracture line found an occult fracture in around three quarters, supporting immobilisation and follow-up.[6] Salter-Harris grades IV and V carry the highest premature-physeal-closure risk and warrant orthopaedic follow-up.[7] Structured, checklist-based chest-radiograph interpretation improves reproducibility and is teachable.[8] The systematic method, the Salter-Harris classification and the decision-rule thresholds are global and are consistent across the ACEM, ATLS, Royal College of Radiologists and ACR frameworks; the local protocol governs the trauma CT-first pathway and the procedural-sedation service.

The chest radiograph — the ABCDE method in detail

The ABCDE method is the workhorse chest-radiograph sequence used in every ED and ICU round. It encodes the same compartments as the longer RIP-ABCDEFGH sequence but is the version most examinable at the viva, because it is rapid, reproducible and exhaustive in five letters. Every letter is read in turn, and no compartment is skipped.[8]

Common ED chest radiograph findings — a pattern library

The Fellowship candidate is expected to recognise the high-yield chest-radiograph patterns at the bedside and state the disposition. The patterns below are the recurring presentations of the breathless, the febrile and the shocked ED patient. [1]

The abdominal radiograph — the systematic approach

The plain abdominal radiograph (AXR) is a low-yield, high-overuse investigation. Its evidence-based indications are narrow — the suspected bowel obstruction, the perforation (an erect chest film is better), the radiopaque foreign body or ingested packet, and the assessment of an indwelling tube or stent. A systematic read prevents the common error of declaring a non-obstructive film obstructive because of normal calibre bowel gas. [1]

The pelvic radiograph — fractures and the hip

The single AP pelvic radiograph is part of the ATLS primary survey of the major-trauma patient and answers two questions in seconds: is there a mechanically unstable pelvic ring injury, and is there a proximal femoral or acetabular fracture? A systematic read of the rings, the arcs and the hip joints prevents the missed sacral impaction fracture and the subtle pubic rami fracture in the elderly faller. [1]

The cervical spine lateral radiograph — the ABCS method

The lateral cervical spine film was once the cornerstone of the trauma clearance; in modern major-trauma practice the adult cervical spine is cleared by CT. The lateral c-spine film remains high-yield exam material and is still used where CT is unavailable or in the paediatric and low-risk patient. The ABCS method — adequacy, bones, cartilage, soft tissue — is the systematic read. [1]

High-yield trial evidence — the validated imaging decision rules

The imaging decision rules are among the most rigorously validated instruments in all of medicine, and the Fellowship candidate is expected to know the rule, the trial that derived it, and the operating characteristics. The Ottawa ankle rule, the Ottawa and Pittsburgh knee rules, the NEXUS and Canadian C-spine rules, and the paediatric fat-pad meta-analysis are the irreducible minimum.[1][2][3][4][5][6]

Additional chest radiograph pearls — the bedside signs

The high-yield signs below are the recurring discriminating findings on the ED chest film. Each converts a descriptive read into a disposition in seconds — the deep sulcus sign to a decompression, the football sign to a laparotomy, the golden S sign to a bronchoscopy. [1]

Additional comparison tables — the discriminating read

The tube-and-line check — the portable film that saves or kills

Every portable chest film on an intubated or line-bearing patient includes a mandatory line check before the lung fields are read — the tube in the bronchus is the never-event that the radiograph is there to catch. A structured check prevents the misplaced feed, the pneumothorax from a central line, and the over-advanced endotracheal tube. [1]

Additional exam pearls — the plain-film viva

Exam pearls

• Read every film in the same sequence — quality, then compartments, then state the finding and the next action. The method is the mark, not the one-liner.
• ABCCS for the limb: adequacy and alignment, bones, cartilage and joints, soft tissue, summary. The two-view minimum is non-negotiable; a single AP film misses the posterior dislocation and the scaphoid.
• Salter-Harris SALTR: I slipped, II above, III lower, IV through all, V rammed. Grades IV and V carry the growth-arrest risk and need orthopaedic follow-up.
• The posterior fat pad is always abnormal — an occult intra-articular elbow fracture in around three quarters of children, even when no line is seen.
• The silhouette sign localises the pathology — an opacity that erases the heart border is lingula or middle lobe; one that erases the diaphragm is lower lobe.
• A skin fold has normal lung markings beyond it; a true pneumothorax does not. The deep sulcus sign on a supine film is a tension pneumothorax until proven otherwise.
• Tubes and lines are part of every chest radiograph — endotracheal tube 2 to 3 cm above the carina, central line tip in the cavoatrial junction, nasogastric tube not in a bronchus.
• A normal radiograph excludes almost nothing — immobilise the suspicious scaphoid and repeat at 7 to 14 days or CT it. [1]

Red flags