Department of Orthopaedic Surgery, Research School CAPHRI, Maastricht University Medical Center, P. Debyelaan 25, 6229 HX, Maastricht, The Netherlands.
Eur Spine J. 2019 Apr;28(4):775-782. doi: 10.1007/s00586-018-05877-z. Epub 2019 Jan 7.
The aim of this study was to develop and internally validate a multivariable model for accurate prediction of surgical site infection (SSI) after instrumented spine surgery using a large cohort of a Western European academic center.
Data of potential predictor variables were collected in 898 adult patients who underwent instrumented posterior fusion of the thoracolumbar spine. We used logistic regression analysis to develop the prediction model for SSI. The ability to discriminate between those who developed SSI and those who did not was quantified as the area under the receiver operating characteristic curve (AUC). Model calibration was evaluated by visual inspection of the calibration plot and by computing the Hosmer and Lemeshow goodness-of-fit test.
Sixty patients (6.7%) were diagnosed with an SSI. After backward stepwise elimination of predictor variables, we formulated a model in which an individual's risk of an SSI can be computed. Age, body mass index, ASA score, degenerative or revision surgery and NSAID use appeared to be independent predictor variables for the risk of SSI. The AUC was 0.72 (95% CI 0.65-0.79), indicating reasonable discriminative ability.
We developed and internally validated a prediction model for SSI after instrumented thoracolumbar spine surgery using predictor variables of standard clinical practice that showed reasonable discriminative ability and calibration. Identification of patients at risk for SSI allows for individualized patient risk assessment with better patient-specific counseling and may accelerate the implementation of multi-disciplinary strategies for reduction of SSI. These slides can be retrieved under Electronic Supplementary Material.
本研究旨在利用西欧一家学术中心的大样本量，开发并内部验证一种用于准确预测器械辅助脊柱手术后手术部位感染（SSI）的多变量模型。
收集了 898 例接受胸腰椎后路器械融合的成年患者的潜在预测变量数据。我们使用逻辑回归分析来建立 SSI 预测模型。通过计算受试者工作特征曲线（ROC）下面积（AUC）来衡量区分发生 SSI 和未发生 SSI 患者的能力。通过校准图的直观检查和 Hosmer 和 Lemeshow 拟合优度检验来评估模型校准。
60 例（6.7%）患者被诊断为 SSI。经过向后逐步消除预测变量，我们制定了一个模型，其中可以计算个体发生 SSI 的风险。年龄、体重指数、ASA 评分、退行性或翻修手术以及 NSAID 使用似乎是 SSI 风险的独立预测变量。AUC 为 0.72（95%CI 0.65-0.79），表明具有合理的判别能力。
我们使用标准临床实践中的预测变量开发并内部验证了一种用于器械辅助胸腰椎手术后 SSI 的预测模型，该模型具有合理的判别能力和校准度。识别发生 SSI 的高风险患者可以进行个体化的患者风险评估，从而进行更具针对性的患者咨询，并可能加速实施多学科策略以降低 SSI 发生率。这些幻灯片可以在电子补充材料中找到。
