Assessing the Clinical Safety Profile of Computer-Assisted Navigation for Posterior Cervical Fusion: A Propensity-Matched Analysis of 30-Day Outcomes.

Background Computer-assisted navigation (CAN) has been shown to improve accuracy of screw placement in procedures involving the posterior cervical spine, but whether the addition of CAN affects complication rates, neurologic or otherwise, is presently unknown. The objective of this study is to determine the effect of spinal CAN on short-term clinical outcomes following posterior cervical fusion. Methods The American College of Surgeons National Surgical Quality Improvement Program database was queried from 2011 to 2018. Patients receiving posterior cervical fusion were identified and separated into CAN and non-CAN cohorts on the basis of a propensity score matching algorithm to select similar patients for comparison. Rates of 30-day unplanned readmission, reoperation, and other complications were evaluated. A separate matching algorithm was used to generate a subgroup of patients undergoing C1-C2 or occiput-C2 fusion for comparison of the same outcomes. Results A total of 12,578 patients met inclusion criteria, of which 689 received CAN and 11,889 did not. After adjusting for baseline differences, patients receiving CAN experienced longer operations and had higher total relative value units associated with care. There were no significant differences in 30-day complication, readmission, or revision rates. At the occipitocervical junction, there were more hardware revisions in the non-CAN group, but this effect did not reach statistical significance (2 vs. 0; P = 0.155). Conclusions Surgeons should embrace navigation in the cervical spine at their own discretion, as use of CAN does not appear to be associated with increased rates of surgical complications or readmissions despite longer operative time.