316 Optimizing Safety in Robotic Lumbar Instrumented Fusions: A Risk Factor Analysis of Robotic Failures

Abstract

INTRODUCTION: There may be early complications with robot-guided lumbar fusions (RGLF). This study determines risk factors for robot-related complications and suggests optimal OR workflow. METHODS: Single-institution retrospective review of 525 RGLFs. Univariate analysis (independent samples t-tests, chi-squared analysis) assessed demographics, Hounsfield units (HU), procedural characteristics, and OR workflow as risk factors for robot registration failures and pedicle screw malpositioning. Multivariate logistic regression of significant factors was conducted. Subanalysis of Anterior Lumbar Interbody Fusion (ALIF) with RGLF was conducted. Receiver operating characteristic (ROC) curve determined thresholds for independent risk factors. RESULTS: All RGLF: Lower L1-HU and L5-S1 inclusion were significant risk factors for registration failure on univariate analysis. Placing pedicle screws directly after CT acquisition (PS after CT) and before interbody placement showed fewer registration failures. On multivariate analysis, L1-HU (Odds Ratio (OR): 1.012; p = 0.006), L5-S1 inclusion (OR: 2.795; p = 0.033) were independent risk factors. PS after CT improved registration (OR: 0.145; p = 0.004). Age and female gender were risk factors for pedicle screw malpositioning on univariate analysis. PS after CT showed fewer pedicle screw malpositions on univariate and multivariate (OR: 0.146; p = 0.006) analysis. ALIF with RGLF: PS after CT had fewer registration failures. Low L1-HU units was a risk for registration failure on univariate and multivariate (OR: 1.022; p = 0.001) analysis. ROC analysis of L1-HU showed a 145.85 cutoff. Age and hyperlordotic interbody use were risk factors for pedicle screw malposition on univariate analysis. CONCLUSION: Early robot-related complications may be avoided by recognizing risks. PS after CT, before interbody placement may avoid complications including pedicle screw malpositioning. Poor bone quality can lead to registration failures.

Ähnliche Arbeiten