Rib fractures are uncommon in infants. Child abuse must be suspected, especially when location is posterior, as explained by the lever phenomenon.¹ The positive predictive value of rib fractures as an indicator of abuse is 95–100%.² Bone fragility diseases, severe cough, and cardiopulmonary resuscitation can cause rib fractures, and chest physical therapy (CPT) has only been mentioned in a recent retrospective series.³

From May 2000 to May 2003 we prospectively collected chest radiographs performed as a workup for bronchiolitis, and collected six cases of infants less than 2 years old for whom lateral rib fractures or sequelae were diagnosed. With assistance from clinics, biology, radiology, and follow up, child abuse was ruled out. CPT was the only aetiology retained. It consisted in repetitive anterior cephalocaudal compressions and provoked cough, following French national consensus.⁴ Twelve of 14 fractures were located on the lateral part of the fourth to seventh ribs, none at the costovertebral junction; physiotherapists hypothesise (unpublished data) that during CPT, maximum pressure is located in the anterior mid-thorax, namely the fifth and sixth ribs, without any lever phenomenon. It is notable that 12/14 lesions consisted of periosteal reactions with no direct signs of fractures; this may relate to the hypothesis that repeated CPT leads to sub-periosteal haemorrhages more than to real fractures.

In conclusion, rib fractures secondary to CPT seem less unusual than initially reported. We are thus thorough in assessing a non-accidental injury.

Paediatricians must consider the devastating psychological effects of a wrong suspicion of child abuse on the entire family. The benefit of CPT in bronchiolitis should be validated.

To assess diagnosis, radiologists must precisely determine the location of the fractures on the chest and along the rib, and precisely describe radiological features.