Editor,—We read with interest Mike Clancy's comprehensive review of clearing the cervical spine in adult trauma victims.¹ It highlights the current diverse methods of treating cervical spinal injury in the UK and USA, the difficulties of confidently excluding an unstable cervical spine injury in unconscious patients, and the problems associated with unnecessary cervical spine immobilisation. We report our initial experience with the use fluoroscopy to dynamically clear the cervical spine in obtunded patients.

Since 1994 the following protocol has been adopted by the trauma service in Oxford for clearing the spine in the unconscious trauma patient. Anteroposterior and lateral radiographs are taken of the cervical, thoracic and lumbar spines, which may include one attempt at a swimmer's or an oblique view if the cervicothoracic junction is not seen. All patients undergo computed tomography of C1 and C2 cervical vertebrae, as well as at C7 and T1 if the cervicothoracic junction is not adequately visualised. The open mouth “peg view” is extremely difficult to achieve in the collared and intubated patient and has been abandoned in our unit. Unless an unstable injury is identified by the above imaging, the cervical spine is then screened dynamically at the earliest convenient opportunity (using a mobile C-arm BV29 Philips Image Intensifier) by a consultant trauma orthopaedic surgeon putting the neck through a progressively increasing range of movement until full flexion and extension has been achieved.

Between April 1994 and October 1997, 78 adult patients underwent dynamic screening of the cervical spine performed at a median of one day (range 0–12) after admission.

Five of these patients (6.4%) had a cervical fracture or instability; fractures in three patients were readily apparent before dynamic screening and this test was used to confirm stability and allow collar removal. One stable spinous process fracture was identified during dynamic screening that had previously been missed. One patient was demonstrated to have had gross atlantoaxial instability in the absence of a fracture, and subsequently underwent surgical internal fixation; complete rupture of the interlaminar ligaments was confirmed.

Dynamic cervical screening was negative in 73 unconscious patients. Of these, 12 died mainly secondary to associated major intracerebral injury at a median of 4.5 days after injury. In the remaining 61, the cervical spine was cleared at a median of three days (range 5–33) before extubation and one day after admission to ITU. A total of 314 “days in collar” were saved over the study period. None of the deaths in the negatively screened patients were attributable to cervical injury and there were no adverse sequelae from screening in the survivors.

Dynamic cervical screening, unlike other imaging methods, has the advantage of providing direct evidence of cervical movement under controlled and increasing stresses. It can detect new injuries as well as evaluate the stability of known or suspected fractures identified by plain radiographs. The procedure can be readily and rapidly performed in the resuscitation room, angiography suite, theatres, or the intensive care unit. The many benefits of early collar removal have been emphasised in Clancy's excellent review.¹

The method does, however, require specific training and experience, and may not be feasible in a small proportion of patients because of their shape and size.

Our findings add support to the observations of others^{2, 3} that dynamic fluoroscopy is safe, sensitive and specific when used as part of the described spinal injury imaging protocol to identify unstable cervical spinal injuries in unconscious patients.

We have since implemented an identical spinal injury imaging protocol for unconscious injured children, the results of which are being currently analysed.