脊柱侧凸后路矫形术后呼吸系统并发症的危险因素分析

Risk factors of postoperative respiratory complication following posterior correction for spinal scoliosis

目的：分析脊柱侧凸后路矫形术后呼吸系统并发症（respiratory complication, RC）的危险因素并构建预测模型，以指导临床防治。方法：收集中山大学附属一院2006年7月-2011年12月行后路矫形术的脊柱侧凸患者共306例，其中男98例，女208例，年龄6～35岁，平均16．3±5．8岁。记录患者术前相关情况，包括性别、年龄、身高、体重、主胸弯程度（主胸弯Cobb角）、病程、肺功能、中性粒细胞计数、中性粒细胞／淋巴细胞比值（neutrophil-to-lymphocyte ratio,NLR）等；术中变量，包括手术方式、手术时间（切皮至缝皮结束）、麻醉时间（诱导至停止麻醉）、带气管导管时间、是否全凭静脉麻醉、输入液体的晶胶比、最低收缩压、血红蛋白（Hb）丢失量、体温、气道峰压、椎弓根螺钉数、融合节段数及应用甲强龙情况等；术后变量，包括术后血制品输入、镇痛方式、镇痛药物选择、术后进ICU、术后谵妄等。采用单因素和多因素Logistic回归分析筛选与发生术后RC相关的危险因素，并构建预测模型。结果：58例（19．0％）患者术后发生RC。单因素分析结果显示患者病程、术前主胸弯Cobb角、术前中性粒细胞计数、术前中性粒细胞／淋巴细胞、术前肺功能、手术方式、麻醉时间、手术时间、带气管导管时间、术中输液晶胶比、失血量、椎弓根螺钉数、融合节段数、全凭静脉麻醉、大剂量甲强龙、术后谵妄、术后进ICU、镇痛方式等因素对术后RC的影响有统计学差异（P〈0．05）。Logistic回归分析提示麻醉时间（x1）、全凭静脉麻醉（X2）、主胸弯Cobb角（X3）、融合节段数（X4）、术后谵妄（X5）和手术失血量（X6）是脊柱侧凸后路矫形术后发生RC的独立危险因素，多元回归模型为P=1／[1＋exp（0．020X1—1．407X2—0．060X3＋0．574X4＋4．023X5＋0．087X6—8．742）1。结论：术前主胸弯Cobb角大、手术融合节段数多、术中失血量大及术后出现谵妄的脊柱侧凸患者后路矫形术后容易发生RC，麻醉时间长、使用全凭静脉麻醉可能增加术后RC的风险，应加强控制。

Objectives： To investigate the risk factors of postoperative respiratory complication（RC） in patients undergoing posterior scoliosis correction and to establish a predictive model as guidance for clinical manage- ment. Methods： The perioperative data of 306 patients undergoing posterior scoliosis correction under general anesthesia from July 2006 to December 2011 in the First Affiliated Hospital of Sun Yat-sen University were collected. There were 98 males and 208 females with an average age of 16.3 years（SD, 5.8 years, range 6-35 years old）. The patents＇ sex, age, height, weight, preoperative main thoracic Cobb angle, course of disease, preoperative pulmonary function, preoperative neutrophil and neutrophil-to-lymphocyte ratio （NLR）, surgery types, surgery time, anesthesia time, duration of intratraeheal duct, total intravenous anesthetics（TIVA）, crystal- loid/eolloidal intraoperatively, lower intraoperative systolic blood pressure（SBP）, intraoperative blood loss, tem- perature, peak airway pressure, number of pedicle screws, fusion segments, use of methylprednisolone, blood transfusion, postoperative analgesic methods and drugs, ICU staying, postoperative delirium and so on wererecorded. The potential perioperative risk factors of postoperative RC were identified and a predictive model of individual RC risk was established using univariate analysis and multivariate Logistic regression. Results： After searching the patient medical record database in our hospital retrospective/y, 306 patients undergoing scoliosis surgery were included in the study. The incidence of postoperative RC was 19.0%（58/306）. Univariate analysis showed statistically significant difference （P〈0.05） between two tested groups with respect to course of disease, preoperative main thoracic Cobb angle, preoperative neutrophil and NLR, preoperative pulmonary function, surgery types, anesthesia time, surgery time, duration with intratracheal duct, crystalloid/colloidal intraoperative- ly, intraoperative blood loss, number of pediele screws, fusion segments, TIVA, high dose of methylpred- nisolone, postoperative delirium, ICU staying, postoperative analgesia methods. Logistic regression showed that the risk factors of postoperative RC included time of anesthesia（X1）, the usage of TIVA（X2）, thoracic Cobb an- gle（X3）, number of fusion segments（X4）, postoperative delirium（X5） and blood loss during surgery（X6）. The pre- dictive equation was P=l/[1＋exp （0.020X1-1.407X2-0.060X3＋0.574X4＋4.023X5＋0.087X6-8.742）]. Conclusions： Patients with a greater thoracic Cobb angle, more fusion segments, more blood loss during scoliosis surgery, and suffering postoperative delirium are more likely to suffer postoperative RC. Longer time of anesthesia and the use of TIVA could increase the risk of RC.