Comparative risk of instrumentation failure in dynamic and rigid spinal stabilization: a propensity-matched cohort study.

This study aims to compare the risk of instrumentation failure between dynamic and rigid posterior spinal stabilization systems using propensity score matched cohorts, and to investigate the role of spinopelvic parameters and patient-related factors in construct longevity. A retrospective analysis was conducted on 952 patients who underwent thoracic, thoracolumbar, thoracolumbosacral, lumbar, or lumbosacral instrumentation between 2018 and 2024. Instrumentation failure (rod fracture, screw fracture, migration, pull-out, or loosening) was identified in a subgroup of 106 patients. Propensity score matching (1:1 nearest-neighbor, caliper 0.1) was applied using age, sex, osteoporosis, number of instrumented segments, and involvement of S1 as the lowest instrumented vertebra, yielding two matched cohorts of 196 dynamic and 196 rigid stabilized cases with improved baseline comparability. Survival differences were assessed by Kaplan-Meier and log-rank tests, and multivariable Cox regression was performed in both the full cohort and matched groups to identify potential risk factors. The overall failure rate was 11.1%, with a significantly higher rate in rigid constructs than in PEEK rod/dynamic screw constructs (22.0% vs. 6.7%, p < 0.001). In the matched cohort, failure remained more frequent in the rigid group than in the dynamic group (24.5% vs. 10.7%, p < 0.001). Kaplan-Meier analysis demonstrated lower mechanical failure-free survival in rigid constructs (log-rank p < 0.001). In the matched Cox regression analysis, rigid stabilization (HR = 2.83, 95% CI 1.41-5.68), osteoporosis (HR = 4.40, 95% CI 2.08-9.27), PI-LL mismatch (HR = 1.03 per degree, 95% CI 1.01-1.06), and the age-number of instrumented segments composite variable (HR = 1.27, 95% CI 1.04-1.54) were associated with increased hazard of mechanical failure. In this large retrospective propensity-matched cohort with long-term follow-up, dynamic stabilization systems were associated with a significantly lower and delayed risk of instrumentation failure compared with rigid constructs. Restoration of sagittal alignment and consideration of patient-specific factors, including bone quality and spinopelvic morphology, appear important in reducing hardware-related complications. However, because of the observational design, residual confounding, and remaining imbalance after matching, these findings should be interpreted as associative and do not establish definitive superiority of one construct over another.