Treated conservatively, hyperextension injuries of the cervical spine have a poor outcome, but surgical treatment does not yet provide a realistic alternative. This study was prompted by the lack of classifications of cervical spondylosis and outcome studies of hyperextension in the literature, and the absence of a user-friendly neurological score.

The retrospective study included 60 patients admitted over the last 12 years with hyperextension injuries and varying degrees of neurological deficit. The mean age of patients was 52 years and most had radiological evidence of cervical spondylosis. None had any neurological deficit before the accident. Injuries were sustained in falls in 56%, in motor vehicle accidents in 34%, in assaults in 8% and in sports injuries in 2%. The neurological deficits varied: 11% had complete lesions, 33% central cord syndrome, 18% motor complete-sensory incomplete, 33% motor incomplete-sensory incomplete. The nervous system was normal in 2%. The posterior columns were often involved, with loss of pro-prioception. In the majority of cases conservative treatment consisted of six weeks of light traction in gentle flexion, followed by mobilisation in an ABCO brace for six weeks.

The results showed that the mean Asia score gain for the group treated conservatively was 23, for the surgically stabilised group 3 and the surgically- decompressed group 55. There was a 16% mortality rate. The mean time for rehabilitation was 5.5 months. Both the final outcome and the time to rehabilitation were extremely variable.

We present a classification of cervical spondylosis and ways of measuring congenital and acquired spinal cord narrowing. We combine the Asia and Frankel scoring systems to give a user-friendly guide.