The detection of significant fractures in suspected infant abuse

Skeletal survey is a commonly used means of detecting fractures in infants, and is used as a screen in suspected cases of physical abuse. It is recognised that in live infants, a repeat survey some days after a suspected episode of injury will detect more fractures than one taken shortly after the suspected injury, indicating that the latter lacks sensitivity. In infants who die soon after a suspected episode of physical abuse, the managing clinicians do not have the option of a second survey; however there is the opportunity for the microscopic examination of bones removed at autopsy. Increasingly Osteoarticular Pathology at the Manchester University NHS Foundation Trust (MFT) is being sent samples of bones from infants suspected of inflicted injury for histological examination, both from bones with fractures detected at autopsy or skeletal survey and from posterior ribs and long bone metaphyses (sites of significance in assessing for abusive injury) when there is no evidence of fracture on skeletal survey or autopsy. Here we report the results of an audit of the anonymised data from a series of such cases, to establish the sensitivity of skeletal survey (SS) to detect fractures and to define the medico-legal value of submitting bones for histological examination.

This was an audit of skeletal injuries in 38 infants aged <18 months presenting to MFT for specialist histopathological evaluation of suspected non-accidental fractures between January 2011 and June 2017. Histopathological examination was performed on all bones submitted and compared with contact radiography of isolated bones and post-mortem skeletal surveys undertaken by specialist paediatric or musculoskeletal radiologists for the presence of fracture.

A total of 318 fractures were detected histologically; of these, 178 (56%) were of the ribs, 119 (37.5%) were of major limb long bones, 10 (3%) were of the skull, and 11 (3.5%) were recorded as 'other'. Excluding refractures, skeletal survey detected 54% of the fractures recorded histologically. No fractures were detected radiologically that were not seen histologically. Generally, for skeletal survey, detection rates improved with the age of the lesion, and rib fractures were more difficult to detect than long bone fractures. Ribs 5-8 were the most frequently fractured ribs, and metaphyses around the knee accounted for most metaphyseal limb long bone fractures undetected by SS.

In infants coming to post-mortem, histopathology is more sensitive than SS for the detection of clinically significant fractures. In children suspected of non-accidental injuries but with negative or equivocal SS, sampling of the anterior and posterior end of ribs 5-8 and the bones around the knee for histological examination could reveal clinically unsuspected fractures and significant evidence of physical abuse. 71% of infants showed evidence of old fractures typical of non-accidental injury.