Wick Joseph B, Blandino Andrew, Smith Justin S, Line Breton G, Lafage Virginie, Lafage Renaud, Kim Han Jo, Passias Peter G, Gum Jeffrey L, Kebaish Khaled M, Eastlack Robert K, Daniels Alan, Mundis Gregory, Hostin Richard, Protopsaltis Themistocles, Hamilton D Kojo, Kelly Michael P, Gupta Munish, Hart Robert A, Schwab Frank J, Burton Douglas C, Ames Christopher P, Lenke Lawrence G, Shaffrey Christopher I, Bess Shay, Klineberg Eric
Department of Orthopedic Surgery, University of California, Davis, Sacramento, CA, USA.
Department of Neurosurgery, Medical Center, University of Virginia, Charlottesville, VA, USA.
Global Spine J. 2025 Mar;15(2):621-632. doi: 10.1177/21925682231202782. Epub 2023 Sep 19.
Retrospective review.
The International Spine Study Group-AO (ISSG-AO) Adult Spinal Deformity (ASD) Complication Classification System was developed to improve classification, reporting, and study of complications among patients undergoing ASD surgery. The ISSG-AO system classifies interventions to address complications by level of invasiveness: grade zero (none); grade 1, mild (e.g., medication change); grade 2, moderate (e.g., ICU admission); grade 3, severe (e.g., reoperation related to surgery of interest). To evaluate the efficacy of the ISSG-AO ASD Complication Classification System, we aimed to compare correlations between postoperative length of stay (LOS) and complication severity as classified by the ISSG-AO ASD and traditional major/minor complication classification systems.
Patients age ≥18 in a multicenter ASD database who sustained in-hospital complications were identified. Complications were classified with the major/minor and ISSG-AO systems and correlated with LOS using an ensemble-based machine learning algorithm (conditional random forest) and a generalized linear mixed model.
490 patients at 19 sites were included. 64.9% of complications were major, and 35.1% were minor. By ISSG-AO classification, 20.4%, 66.1%, 6.7%, and 6.7% were grades 0-3, respectively. ISSG-AO complication grading demonstrated significant correlation with LOS, whereas major/minor complication classification demonstrated inverse correlation with LOS. In conditional random forest analysis, ISSG-AO classification had the greatest relative importance when assessing correlations across multiple variables with LOS.
The ISSG-AO system may help identify specific complications associated with prolonged LOS. Targeted interventions to avoid or reduce these complications may improve ASD surgical quality and resource utilization.
回顾性研究。
国际脊柱研究组- AO(ISSG-AO)成人脊柱畸形(ASD)并发症分类系统旨在改善接受ASD手术患者并发症的分类、报告及研究。ISSG-AO系统根据侵入程度对处理并发症的干预措施进行分类:0级(无);1级,轻度(如药物调整);2级,中度(如入住重症监护病房);3级,重度(如与相关手术有关的再次手术)。为评估ISSG-AO ASD并发症分类系统的有效性,我们旨在比较ISSG-AO ASD分类系统和传统的主要/次要并发症分类系统对术后住院时间(LOS)和并发症严重程度的相关性。
在多中心ASD数据库中识别年龄≥18岁且发生院内并发症的患者。使用主要/次要分类系统和ISSG-AO系统对并发症进行分类,并使用基于集成的机器学习算法(条件随机森林)和广义线性混合模型将其与LOS进行相关性分析。
纳入了19个研究点的490例患者。64.9%的并发症为主要并发症,35.1%为次要并发症。根据ISSG-AO分类,0-3级分别为20.4%、66.1%、6.7%和6.7%。ISSG-AO并发症分级与LOS呈显著相关性,而主要/次要并发症分类与LOS呈负相关。在条件随机森林分析中,ISSG-AO分类在评估多个变量与LOS的相关性时具有最大的相对重要性。
ISSG-AO系统可能有助于识别与住院时间延长相关的特定并发症。避免或减少这些并发症的针对性干预措施可能会提高ASD手术质量和资源利用效率。