Two- versus multi-rod constructs for adult spinal deformity: A systematic review and Random-effects and Bayesian meta-analysis.

• Multiple-rod constructs associated with reduction in rod fracture, pseudoarthrosis, and re-operation when compared with 2-rod constructs. • Country of primary study was found to be significant modulator of heterogeneity for rod fracture. • Similar rates observed of proximal junctional kyphosis between multiple- and 2-rod constructs. Surgical approaches in adult spinal deformity are associated with high rates of adverse events including hardware failure and rod fracture. Recently, some reports have emerged comparing multiple-rod constructs with 2-rod constructs suggesting potential benefits with the former. However, these have been limited by variability in observed outcomes, which have limited the change of paradigm in adult spinal deformity surgery. To compare the rate of rod fracture, pseudoarthrosis, proximal junctional kyphosis and re-operation between 2-RC and M-RC. MEDLINE/Pubmed, Web of Science and Embase were searched without language restrictions for relevant articles from inception until October 2021. All observational cohort studies assessing patients with ADS undergoing 3-column osteotomy and comparing 2-RC with M-RC procedures on pseudarthrosis, rod fracture, kyphosis or reoperation were included. Data were independently extracted by 2 authors. Random-effects and Bayesian meta -analysis were used. Six primary studies met inclusion criteria, yielding a total of 448 participants, with 223 receiving 2-RC and 225 M-RC. The random-effects meta -analysis pointed to a significantly lower risk of rod fracture associated with M-RC (RR = 0.43, 95 %CI = 0.28–0.66), with moderate heterogeneity being observed (I² = 20 %, p = 0.28). The random-effects meta -analysis pointed to a lower risk of pseudoarthrosis with M-RC than with 2-RC (RR = 0.49, CI = 0.28–0.84, to a lower rate of re-operation with M-RC than with 2-RC (RR = 0.52, CI = 0.28–0.97) and to a similar rate of proximal junctional kyphosis between 2-RC and M-RC patients (RR = 0.91, CI = 0.60–1.39). Low heterogeneity was observed for studies comparing pseudoarthrosis (I² = 9 %, p = 0.35), re-operation (I² = 0 %, p = 0.41) and proximal junctional kyphosis (I² = 0 %, p = 0.85). These findings suggest that multiple rod-fracture constructs are associated with lower rates of rod fracture, re-operation rates, pseudoarthrosis but not proximal junctional kyphosis. Future studies should address the impact of other modulators of heterogeneity such as body mass index, metal alloys and length of the constructs. [ABSTRACT FROM AUTHOR]