Multivariate analysis of the neurological outcome of surgery for cervical compressive myelopathy

The neurological outcome of decompressive surgery for cervical myelopathy is influenced by several factors. Although each factor may have an independent effect, it is more likely that the outcome is influenced by more than one factor. We examined the results of multivariate analysis and multiple regression analysis of the neurological outcome of patients treated by cervical cord decompression.

A total of 77 patients with cervical spondylotic myelopathy (43 men, 34 women) and 58 with ossification of the posterior longitudinal ligament (OPLL) (39 men, 19 women) were studied with an average follow-up interval of 8.3 years. The clinical data, neurological and radiological findings, and results of spinal cord evoked potentials (SCEPs) were retrieved from the medical records and included in the analysis.

Multivariate analysis indicated that the outcome for patients with spondylosis was positively influenced, in order of importance, by increased transverse area of the cord >or=60%, presence of single-level anterior fusion, a high preoperative neurological score, normal epidural SCEPs, and clinical features of brachialgia and cord type. In patients with OPLL, multivariate analysis showed that the long-term outcome was positively influenced, in order of importance, by the presence of mixed or localized OPLL, normal epidural SCEPs, high preoperative neurological score, a single-vertebra spondylectomy with anterior fusion, laminoplasty, widening of the transverse area of the cord >or=40%, and an expansion rate of the spinal canal after laminoplasty >or=40%.

We suggest that multivariate analysis is useful for assessing the neurosurgical outcome in patients with cervical compressive myelopathy.