Postoperative complication rates and hazards-model survival analysis of revision surgery following occipitocervical and atlanto-axial fusion

• Main Authors: Daniel S. Yang, Shyam A. Patel, Kevin J. DiSilvestro, Neill Y. Li, Alan H. Daniels
• Format: Article
• Language: English
• Published: Elsevier 2020-10-01
• Series: North American Spine Society Journal
• Subjects: Occipitocervical; Atlanto-axial; Fusion; Complications; Revision; Survival analysis
• Online Access: http://www.sciencedirect.com/science/article/pii/S2666548420300172

Background: Complication rates following occipitocervical and atlanto-axial fusion are high. While methods to fuse the upper cervical spine levels have evolved, complication rates and surgical survivorship of occipitocervical fusion versus atlanto-axial fusion are incompletely understood. Methods: The PearlDiver Research Program (www.pearldiverinc.com) was used to identify patients undergoing primary occipitocervical or atlanto-axial fusion between 2007 and 2017. Incidence of each fusion procedure was studied across time. Multivariable logistic regression was used to compare 30-day readmission, 30-day medical complications, and post-operative opioid utilization at 1, 3, 6, and 12 months between cohorts, controlling for age, gender, Charlson Comorbidity Index (CCI), and indication for surgery. Risk of revision was compared through Cox-proportional hazards modeling, Kaplan-Meier survival, and log-rank test. Results: Cohorts of 483 occipitocervical fusions and 737 atlanto-axial fusions were examined. From 2008 to 2016, incidence of occipitocervical fusion rose 55.9%, whereas atlanto-axial fusion rose 21.6%. A greater percentage of atlanto-axial fusions were due to trauma (69.9% vs. 50.5%), whereas a greater percentage of occipitocervical fusions were due to degenerative disease (41.6% vs. 29.4%) (p = 0.0161). Total 30-day complications were seen in 40.9% of occipitocervical fusion patients compared to 26.3% of atlanto-axial fusion patients (aOR=2.06, p < 0.0001). Risk of surgical site infection was increased (aOR=2.59, p = 0.0075). Kaplan Meier survival analysis and Cox-proportional hazards demonstrated greater risk of revision following surgery for occipitocervical fusion (log rank: p < 0.0001, aHR=2.66, 95%CI 1.73–4.10, p < 0.0001). Conclusions: Rates of occipitocervical and atlanto-axial fusion are rising, while complication and revision surgery rates remain high, with occipiticervical fusion leading to higher rates even after controlling for patient characteristics and surgical indication. Spine surgeons should be cautious when considering fusion of the occipitocervical levels if atlanto-axial fusion could be performed safely and provide adequate stabilization to treat the same pathology.