Spine. 2026;51(20):E519-E529
The extent and complexity of deformity correction deserve explicit consideration when discussing perioperative risk. Katayanagi and colleagues identify associations between operative characteristics and complications during the first 14 days, but their registry does not establish that changing approach or staging surgery will independently prevent harm. Understanding the outcome definition is essential before applying its headline complication rate.
Study and key findings
This retrospective analysis used prospectively recorded Japanese registry data from 3,193 patients undergoing fusion of at least three vertebral segments between 2018 and 2022. Mean age was 60.8 years, and approximately three-quarters underwent fusion of six or more segments.
At least one complication occurred in 910 patients (28.5%); 812 (25.4%) had a major complication, and three died. Importantly, blood loss exceeding 1,000 ml was classified as a major complication and occurred in 638 patients (20.0%). Neurological complications occurred in 3.1%.
Adjusted odds of any complication were higher with operative duration exceeding five hours (odds ratio 4.40), three-column osteotomy (3.50), fusion exceeding five levels (1.51), and posterior/transforaminal interbody fusion (1.47). Staged surgery (0.24), lateral lumbar interbody fusion (0.74), and minimally invasive techniques (0.65) were associated with lower odds. Associations were broadly consistent in the long-fusion subgroup.
Interpretation and limitations
These findings support reviewing surgical burden and possible alternatives, not ranking techniques independently of indication. Prolonged surgery may reflect deformity complexity or intraoperative events; five hours is not a validated deadline for completing an operation.
The registry lacked radiographic deformity severity, achieved correction, patient-reported outcomes and several important comorbidities. Residual confounding therefore limits comparisons, particularly the apparently favorable association with staging. The bleeding-based definition substantially influences the composite endpoint, while overlapping major and minor events mean their percentages should not be added. Fourteen-day surveillance cannot characterize delayed infection, pseudarthrosis or later mechanical failure.
Clinical takeaway
Use these data to inform counseling and multidisciplinary planning around anticipated blood loss, fusion extent and osteotomy requirements. Consider staging or less invasive options when anatomically and clinically appropriate, while preserving necessary correction and decompression. Decisions require balancing early procedural burden against durable alignment and function, outcomes this study did not measure.