Guidelines map imaging strategy to address cases of child abuse

Radiological evidence can be vital in identifying children who are suffering physical abuse. But low-quality images and inappropriate studies can severely hamper the accuracy of any such claims. Subtle signs of harm that should be detectable by expert eyes may instead be missed.

One problem is the lack of a standardized approach to imaging this patient population, according to Dr. Karl Johnson, a pediatric radiologist at Birmingham Children's Hospital in the U.K. Procedures and protocols vary widely among hospitals. Given the implications of either making a false accusation or overlooking genuine child abuse, many radiology staff members are reluctant to get involved.

In a bid to address this problem and reinforce radiologists' responsibility for child protection, the U.K. Royal College of Radiologists (RCR) has published a comprehensive guidance document. "Standards for Radiological Investigations of Suspected Non- Accidental Injury," produced jointly with the U.K. Royal College of Pediatrics and Child Health, includes the latest recommendations on imaging protocols, systems for image review, reporting procedures, and radiation safety issues.

"Part of this document is controversial; part of it is aspirational," Johnson, said at the 2008 U.K. Radiological Congress (UKRC), held in Birmingham. "We fully accept these guidelines may need to be rewritten in three or four years' time as we learn from people's experiences."

Poor communication between pediatric departments and radiology departments is currently the weakest link in the chain of investigating suspected child abuse, he said. The dialogue has to be a two-way process, with senior staff taking a leading role.

The RCR guidelines recommend that a full skeletal survey be performed for all children under two years old for whom nonaccidental injury (NAI) is one of the possible diagnoses. This survey is used to detect occult bony injuries, obtain further information about a clinically suspected injury, aid in dating bone injuries, and help diagnose any underlying disorder that may predispose the child to fractures. If the x-ray survey is not performed, then the reasons for this decision must be documented.

Children over the age of two may also be referred for a full skeletal survey, though this will depend on the social history and clinical and physical findings.

The skeletal survey should ideally be performed within 24 hours of referral during the normal working day, Johnson said. Scanning may be delayed if the child is unstable or may be performed out of hours in urgent cases.

Radiology staff should be told why they are performing the examination so it can be tailored appropriately. Information relevant to the child's development, such as prematurity or a family history of nutritional rickets, should also be passed on, as should the severity of any incident that has been used to explain an injury.

The RCR recommends that two radiographers should perform the skeletal survey, while another healthcare professional should be present in the radiology department to take responsibility for the child's safety.

Although it may be necessary to hold the child during x-ray acquisition, care must be taken not to obscure pathological findings, said Dr. Amaka Offiah, a radiologist at Great Ormond Street Hospital for Children in London, speaking at the same UKRC session. "We did an audit a few years ago where we looked at 50 consecutive skeletal surveys referred to us," she said. "We found that 35% of the radiographs had artifacts, and most of these were the assistant's hands."

The skeletal survey advocated by the RCR comprises 20 separate acquisitions. These include oblique views of both sides of the chest to show the ribs, skull x-rays, and views of both upper and lower extremities: upper arms, forearms, femurs, lower legs, hands, and feet. Single-film "babygrams" should not be performed under any circumstances.

These are not sufficient for a differential diagnosis.

The x-rays should be reviewed immediately by a radiologist to check whether the quality is sufficient for diagnosis or whether some views should be repeated while the child is still in the department. Some equivocal findings may not be resolved immediately. For example, a suspect fracture may be confirmed only by repeating the x-ray after a couple of weeks to check for signs of healing.

"Not every child should have a repeat skeletal survey; not every child should have repeat views. There should be a discussion about whether this is warranted in each particular case," Johnson said.

Additional CT or MRI studies may be carried out as well, particularly if brain injury is suspected. The RCR recommends that all children under the age of one should have a head CT scan when there is evidence of physical abuse, even if there are no signs of neurological damage. This advice may be controversial, but it is entirely justifiable, he said.

"It is important to remember that it is the child with the occult head injury who is sent home that comes back dead," Johnson said.

When reporting soft-copy images, 1K monitors will probably be sufficient, but if monitors with higher resolution are available, then these should be used, Offiah said. Reporting radiologists should also take advantage of any image manipulation tools on the Volume-rendered cranial postmortem CT in case of suspected atlantoaxial rotatory subluxation. Note position of right lateral mass of first cervical vertebra compared with column of right lateral masses of lower vertebrae. This is a common nonaccidental finding in many postmortem CT examinations. (Provided by Dr. A. Persson, Linköping University Hospital, Sweden) workstation.

"We have all our radiographs double- reported by two and sometimes by three consultants. We have regular meetings with pathologists to discuss the findings that we miss, and we also have support from the various colleges, so radiologists and radiographers have confidence that they will be supported if they make an erroneous diagnosis," she said.

Radiology departments may also play a greater role in postmortem investigations of suspected NAI in the future. An ongoing study at Great Ormond Street, funded by the U.K. Department of Health, is examining the use of whole-body MRI and CT in all pediatric patients. Preliminary results indicate that MRI may be beneficial at spotting genetic metaphyseal disorders, and CT for detecting occult rib fractures, Offiah said.

-By Paula Gould