One- versus Three-- Level Fusion in Patients Undergoing Multilevel Lumbar Decompression: Relative Perioperative Risks and Five-year Revisions.

ABSTRACT

STUDY DESIGN: Retrospective cohort study. OBJECTIVE: To assess relative odds of perioperative complications, readmissions, and five-year survival to reoperation for three-level lumbar decompression patients who undergo three-level fusion relative to one-level fusion. BACKGROUND: Patients undergoing multilevel lumbar decompression may be indicated for fusion at one or more levels. The question of fusing only one level with indications such as spondylolisthesis or fusing all levels decompressed is of clinical interest in both the short and longer term. METHODS: Patients undergoing three-level lumbar decompression were extracted from the PearlDiver M165Orto database. The subset of these patients undergoing concomitant three-level and one-level lumbar fusion were identified and matched 11 based on patient age, sex, and Elixhauser Comorbidity Index (ECI) scores. The incidence and odds of 90-day postoperative adverse events were compared between the two groups by multivariable analysis, and comparative five-year survival to lumbar spinal reoperation was determined. RESULTS: After matching, 28,276 patients were identified as undergoing three-level lumbar decompression with three-level fusion and the same for those undergoing three-level decompression with one-level fusion. Controlling for patient age, sex, and ECI, three-level fusion patients had significantly greater odds ratio (OR) of many 90-day adverse events and aggregated any (OR 1.42), serious (OR 1.44), and minor (OR 1.42) adverse events, as well as readmissions (OR 1.51) (P<0.0001 for all). Five-year survival to reoperation was significantly lower for those undergoing three-level decompression with three-level fusion (P<0.0001). CONCLUSIONS: Three-level lumbar decompression patients who underwent three-level fusion were found to be at significantly greater odds of 90-day postoperative adverse events, readmissions, and five-year reoperations relative to those undergoing one-level fusion. The current data support the concept of limiting fusion to the levels with specific indications in the setting of multi-level lumbar decompressions and not needing to match the decompression and fusion levels.