Non-accidental injury: the radiology of suspected physical abuse

Written by the X-Ray Atlas editorial team from the sources listed below · Medical review pending · Updated 24 October 2026 · Level: Advanced

Non-accidental injury (NAI), or physical abuse, is a leading cause of serious injury and death in infants. Radiographs often provide the first objective evidence, and some fracture patterns are strongly associated with abuse. But an X-ray never diagnoses abuse on its own. The radiologist describes the injuries, gives a range of possible ages and comments on how well they fit the history. The diagnosis is reached by a multi-agency safeguarding process. This page teaches how to recognise concerning findings, how the skeletal survey works in the UK, and, most importantly, what to do when you are worried.

1. When to think of NAI

Consider abuse whenever an injury does not fit the child. The questions to ask of every paediatric fracture are:

• Is the child mobile? Any fracture in a non-ambulant infant (not yet crawling, cruising or walking) needs a safeguarding assessment. A widely quoted phrase is "those who don't cruise rarely bruise", and the same logic applies to fractures.
• Does the mechanism match the injury? A short fall from a sofa rarely causes multiple or complex fractures.
• Is the history consistent? A history that changes, is vague, or is absent for a significant injury is concerning.
• Was there a delay in seeking help?
• Are there other injuries? Bruising in a non-mobile baby is itself a safeguarding concern under NICE CG89.

Thinking of abuse is not the same as accusing anyone. The duty is to recognise the possibility and follow local safeguarding procedures, so that the child is assessed properly.

2. Fracture specificity

Fractures differ in how strongly they suggest abuse. The table follows the widely used scheme from Kleinman's work, as reflected in the RCR and Society and College of Radiographers (SCoR) guidance. Specificity is highest in infants; any fracture becomes more concerning when the history does not fit.

Specificity	Fracture types
High	Classic metaphyseal lesions (corner and bucket-handle fractures); posterior rib fractures; scapular fractures; spinous process fractures; sternal fractures.
Moderate	Multiple fractures, especially bilateral; fractures of different ages; epiphyseal separations; vertebral body fractures and subluxations; finger fractures in a non-ambulant child; complex skull fractures.
Low (common in accidents)	Subperiosteal new bone formation; clavicle fractures; long bone shaft fractures; linear skull fractures.

3. Classic metaphyseal lesions

A classic metaphyseal lesion (CML) is a planar fracture across the immature metaphysis, close to the growth plate. It is thought to result from shaking, twisting or pulling forces on a limb, and is highly specific for abuse in infants.

• Seen side on, the fragment at the edge looks like a small triangle: a corner fracture.
• Seen at an angle, the whole disc of bone looks like a curved rim: a bucket-handle fracture.
• Common sites are the distal femur, proximal and distal tibia, and proximal humerus.
• CMLs are subtle. They need well-collimated, high-quality views of each joint region, which is why the survey uses separate images rather than one whole-body film.
• They often heal with little periosteal reaction, so they can be hard to date.

A CML is not a Salter-Harris fracture, although a type II injury can look similar. Normal variants (such as metaphyseal "spurs" and step-offs) also mimic CMLs, which is one reason the survey should be reported by a radiologist with paediatric experience.

4. Rib fractures

• Posterior rib fractures near the costovertebral junction are strongly associated with abuse. They are thought to result from forceful squeezing of the chest, which levers the rib over the transverse process.
• Acute rib fractures are often invisible. They become visible as callus forms, which is a major reason for the follow-up survey.
• Healing fractures show as rounded areas of callus along the rib, sometimes in a vertical line across several adjacent ribs.
• Cardiopulmonary resuscitation in infants rarely causes posterior rib fractures, and birth trauma rarely causes them. Neither should be assumed without expert review.

Oblique views of the ribs are part of the standard survey because they increase detection. See also chest trauma.

5. The skeletal survey

In the UK, imaging of suspected physical abuse follows the RCR/SCoR guidance The radiological investigation of suspected physical abuse in children (2017, revised 2018). Key principles:

• A skeletal survey is a defined set of separate radiographs of the whole skeleton, including views of the skull, chest (with obliques of the ribs), spine, pelvis and each limb segment, with views of the hands and feet. A single "babygram" is not acceptable.
• It is generally indicated in children under 2 years with suspected physical abuse, and selectively in older children, as directed by the safeguarding paediatrician.
• The examination should be justified and performed by experienced staff, with a chaperone, and reported by a radiologist with appropriate expertise. Many centres recommend two reporters.
• A follow-up survey at about 11–14 days (a limited set of views, according to the guidance) improves detection of rib fractures and CMLs and helps with dating.
• CT head is recommended in children under 1 year with suspected abuse, and in older children where indicated. MRI of the brain and spine may be added.
• Siblings or other children in the household may need assessment, as directed by the safeguarding team.

6. Dating fractures: be cautious

Estimating the age of a fracture can help test a history, but the evidence is limited and healing rates vary with the child's age, the bone and the degree of immobilisation. A rough sequence is:

• Soft-tissue swelling settles over days.
• Periosteal new bone usually appears at about 7–14 days.
• Soft callus at about 2–3 weeks; hard callus later; remodelling over months.

These ranges vary greatly between studies. Reports should give broad ranges and say what timing an appearance is "consistent with", rather than a precise date. Avoid giving dates yourself; leave this to the reporting radiologist and the safeguarding team.

7. Differential diagnoses

Medical causes of fractures or bone changes must be considered, because a wrong label of abuse also causes great harm.

Condition	Clues
Osteogenesis imperfecta	Osteopenia, wormian bones in the skull, bowing, blue sclerae in some types, family history.
Rickets	Cupped, frayed, widened metaphyses and osteopenia (see metabolic bone disease).
Birth trauma	Clavicle and humeral fractures in the newborn period; posterior rib fractures are rare.
Physiological periosteal reaction	Thin, smooth and symmetrical along long bone shafts in infants aged about 1–4 months.
Rarer conditions	Menkes disease, scurvy, congenital syphilis, prematurity-related bone disease.
Accidental injury	A credible, consistent history matching the injury and the child's development; for example a toddler's fracture in a walking child (see children's fractures).

The presence of a bone disease does not exclude abuse, and the two can coexist.

Red flags: act now

What to focus on, by role

Role	Priorities
Medical student	Know the high-specificity fractures and the principle that fractures in non-mobile infants need a safeguarding assessment.
Nurse	Notice and record bruising or injuries in non-mobile babies, observe interactions, and escalate concerns through your safeguarding lead.
Junior doctor	Take and document a careful history, recognise when the injury does not fit, escalate to the senior and safeguarding paediatrician, and do not attempt to date fractures yourself.
Radiographer	Perform the survey to the RCR/SCoR protocol with appropriate staffing, chaperoning and documentation, and ensure the follow-up survey is booked.
Advanced / FRCR	CMLs and their mimics, subtle rib fractures, careful dating language, and reporting in a way that will stand up in court.

References

1. Royal College of Radiologists and Society and College of Radiographers. The radiological investigation of suspected physical abuse in children. 2017, revised 2018. rcr.ac.uk
2. Royal College of Paediatrics and Child Health. Child Protection Evidence: systematic reviews. childprotection.rcpch.ac.uk
3. National Institute for Health and Care Excellence. Child maltreatment: when to suspect maltreatment in under 18s (CG89). nice.org.uk
4. Kleinman PK (ed.). Diagnostic Imaging of Child Abuse. 3rd ed. Cambridge University Press; 2015.
5. Sugar NF, Taylor JA, Feldman KW. Bruises in infants and toddlers: those who don't cruise rarely bruise. Arch Pediatr Adolesc Med 1999;153(4):399–403.

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.