Fractures are a manifestation of physical abuse and common accidental injuries. Distinguishing which fractures are indicative of abuse and optimizing the identification of occult fractures are the challenges.Objectives:
To identify additional studies published since our original systematic reviews to address these two issues.Methods:
An all-language literature search of 14 databases was conducted for the years 2005–2013, using revised keywords. All studies underwent standardized critical appraisal by two independent reviewers, applying quality criteria relating to the confirmation of child abuse, exclusion of abuse and quality of skeletal survey conducted. A meta-analysis, stratified by age, was conducted to determine the predictive value for abuse of specific fractures by fitting a random effects model.Results:
Twenty-three studies addressed ‘radiological investigations’, and nine studies ‘fractures indicative of abuse’. Radiological studies reiterated that a single investigation (skeletal survey or radionuclide imaging) will miss some abusive fractures; in 8.4–37.6% of children, the repeat skeletal survey added new information that influenced the child protection procedures. Debate continues as to the optimal images to include in the repeat skeletal survey. A meta-analysis of femoral and humeral fractures by age highlighted that children younger than 18 months are significantly more likely to have sustained their fracture as a consequence of abuse, than those aged 1–4 years.Authors' Conclusions:
Recent literature validates the original conclusions that repeat skeletal imaging adds important information on fractures. Comparative studies of femoral, humeral, rib and skull fractures enabled a meta-analysis by age, however further comparative studies are needed.