Early complications related to approach in cervical spine surgery: single-center prospective study

Surgical intervention is performed on the cervical spine in a heterogeneous number of pathologic conditions in a diverse patient population. Several authors have examined complication prevalence in cervical spine surgery using retrospective analysis. However, few prospective studies have directly examined perioperative complications. Most prospective studies in the spine literature have assessed only specific spinal implants in carefully selected surgical patients, and complication incidence in broader patient populations is limited.

To prospectively collect data on all patients who underwent cervical spine surgery at a large tertiary care center and to evaluate the effect of the approach and the incidence of early complications.

Data were collected prospectively on 119 patients admitted to the neurosurgical service at Thomas Jefferson University hospital from May to December 2008. Data collected consisted of preoperative diagnosis, medical comorbidities, body mass index, surgical approach, length of stay, and complications, and were analyzed using multivariate regression analysis. Complications occurring within 30 days after each operative procedure were included. Medical adverse events, regardless of their relationship to the operative intervention, were also included as complications. A previously validated binary definition of major and minor complications was used to stratify the data.

Overall, 53 of 119 patients (44.5%) experienced at least one complication. Eleven of 41 patients (26.8%) undergoing only an anterior cervical procedure had a perioperative complication, compared with 26 of 53 patients (49.0%) undergoing only a posterior cervical procedure (P = .01). In patients undergoing a combined anterior and posterior surgical procedure, 16 of 25 (66%) experienced a complication, a significant difference in comparison with solitary anterior procedures (P = .004). Anterior procedures were associated with postoperative dysphagia and vocal cord paresis, whereas wound infection and C5 palsy was more frequently recorded in the group undergoing surgery via an isolated posterior approach.

The incidence of complications or adverse events is not definitely known for most spinal procedures because of the complexity of defining complications and obtaining accurate data. Therefore, to obtain a more accurate assessment of spinal procedures, a prospective algorithm was designed to collect and record complications during the acute perioperative period. Using this technique, a significantly higher complication rate was documented than had been previously reported for cervical spine operative interventions. In addition, use of a broad definition of perioperative complications likely increased the recorded incidence of perioperative adverse events and complications. Complications were more common in patients undergoing posterior and anteroposterior procedures.