Decompression sickness in fish farm workers: a new occupational hazard.

Subjects, methods, and results I assessed 22 infants and children with fractures (mean age 6 months) who were the subjects of proceedings for child protection (19) or whose parents had been charged with causing their injuries (three). I scrutinised the medical notes, court papers, and x ray films for evidence of metabolic bone disease, particularly osteogenesis imperfecta (using the classification of Sillence et al3), copper deficiency,4 metabolic disease of prematurity, and "temporary brittle bones." Fourteen children had at least one rib fracture, which was associated with limb fractures in 10 and with limb and skull fractures in one. Sixteen children had rib fractures or metaphyseal fractures, or both, which are characteristic of non-accidental injury. Absence of bruising at the site of the fracture in 10 children was claimed to be evidence for metabolic bone disease, but nine had clear additional evidence of non-accidental injury. The diagnoses suggested as alternatives to non-accidental injury were osteogenesis imperfecta in nine children, prematurity in two, and copper deficiency in two. In eight children osteogenesis imperfecta, copper deficiency, and temporary brittle bone disease were all offered as alternative diagnoses. In two patients, who were siblings, type I osteogenesis imperfecta had been diagnosed by the paediatrician. No other child had any first order features ofosteogenesis imperfecta-namely, recurrent unexplained fractures while in foster care, blue sclerae in both the child and a parent, a history of recurrent fractures in one parent, multiple wormian bones, or dentinogenesis imperfecta. Second order features of osteogenesis imperfecta, such as joint hypermobility, or discoloured sclerae in a child under 6 months, were reported in several cases but were either unconfirmed or not diagnostic in the absence of first order features. Apart from the fractures, the bones were radiologically normal in all except the two children with type I osteogenesis imperfecta. No evidence of copper deficiency, such as deficient copper intake associated with extreme prematurity or malnutrition, or of osteopenia, delayed bone age, refractory hypochromic anaemia, or neutropenia was found in any of the 10 children in whom it had been suggested. There were two premature infants. Neither had been fed parenterally and neither showed evidence ofosteopenia or rickets. Temporary brittle bone disease was proposed as a diagnosis for seven children. The fractures ceased immediately the children were placed in foster care. All had additional evidence of abuse. Additional evidence of abuse was present in 17 cases (table). In 20 cases the courts decided that the fractures were due to child abuse; in the remaining two, in which the fractures were due to proved osteogenesis imperfecta type I, there was other evidence of abuse. The children were placed in foster care and only one further fracture occurred, four years after placement, when the child fell off a wall.