Best Evidence Topics
  • Send this BET as an Email
  • Make a Comment on this BET

Proximal Humeral Fractures and non-accidental injury in children

Three Part Question

In [children in whom non-accidental injury is suspected] what [is the specificity of] an [isolated proximal humeral fracture]?

Clinical Scenario

A 22 month old child attends the department and is reported not to be using her arm. You question the mother of the child who explains the child seems to have injured herself at nursery, though no one can corroborate the story. X-ray reveals an undisplaced humeral fracture. Further inquiry is unrewarding, and you wonder whether the presence of the fracture alone is sufficient to support a diagnosis of non-accidental injury.

Search Strategy

Medline 1969-10/04 using the OVID interface.
[Exp battered child syndrome OR battered child OR exp child abuse OR non-accidental OR] AND [exp humeral fractures OR {(exp fractures OR fracture$.mp OR exp fractures, closed OR exp fractures, open) AND (exp humerus OR humer$.mp)}]

Search Outcome

Altogether 44 papers were found of which 42 did not address the question directly. The remaining 2 papers are shown below.

Relevant Paper(s)

Author, date and country Patient group Study type (level of evidence) Outcomes Key results Study Weaknesses
Strait RT et al,
124 children under 36 months with humeral shaft fractures identified by retrospective chart review. Subdivided into age less than 15 months and those aged 15 to 36 months. Diagnosis of abuse, indeterminate, or not abuse made by consensus.Diagnostic retrospective test study.Abuse diagnosed - overall<15 months: 36%; 15-36 months:1%Small numbers Single hospital "Gold standard" of diagnosis of abuse inadequate - based on retrospective chart review, with no standard criteria for the definition of abuse "Abuse"/ "not abuse" not determined in 23 cases. If all these were "abuse", figures alter dramatically. In case of Salter Harris I and II and lateral condylar group only single fracture of each type, skewing the validity of the calculations.
Supracondylar fractures<15 months: 2/10 (20%)
Spiral/oblique fractures<15 months: 7/12 (58%)
Shaw BA et al,
34 unselected children (<3 years) with humeral shaft fracturesRetrospective diagnostic test studyTransverseSpec 57% (57%) LR+ 0.39 (0.72)Retrospective Gold standard problem - diagnosis of abuse based on retrospective case-note review and information from child protection services; no standard criteria for the definition of abuse Small numbers Single hospital
ObliqueSpec 79% (71%) LR+ 0.78 (0.27)
SpiralSpec 68% (76%) LR+ 2.07 (2.58)


The detection and diagnosis of child abuse is difficult and requires a holistic approach. However, traditional teaching has suggested that certain injuries are pathognomic of abuse. These papers attempt to define the specificity for child abuse of various types of humeral fracture. However, both studies lack an adequate "gold standard" that abuse was the cause. It is possible that abused children may have been missed and also the converse could be true. Despite this, the reported incidence of abuse associated with these fractures appears high. In particular spiral or oblique fractures, in children under the age of 3 more often occur as a result of abuse.

Clinical Bottom Line

Although not pathognomonic of child abuse, the presence of a humeral fracture in a young child should lead to further investigation of its cause.

Level of Evidence

Level 3 - Small numbers of small studies or great heterogeneity or very different population.


  1. Strait RT, Siegel RM, Shapiro RA. Humeral fractures without obvious etiologies in children less than 3 years of age: when is it abuse? Paediatrics 1995;96(4 Pt 1):677-71.
  2. Shaw BA, Murphy KM, Shaw A, et al. Humerus shaft fractures in young children: accident or abuse? J Pediatr Orthop 1997:17(3):293-7.