胸腰椎爆裂性骨折患者延长住院时间危险因素分析

Analysis of risk factors for prolonged length of stay of patients with thoracolumbar burst fractures

目的 分析胸腰椎爆裂性骨折患者延长住院时间 (length of stay,LOS) 危险因素。方法 回顾分析 2019 年 3 月至 2023 年 6 月我院行长节段椎弓根内固定术治疗的 224 例胸腰椎爆裂性骨折患者术前临床资料。LOS 大于或等于第 75 百分位数的时间为延长 LOS。采用多因素 Logistic 回归模型分析胸腰椎爆裂性骨折患者延长 LOS 的独立危险因素,采用联合接收者操作特征 (receiver operating characteristic,ROC) 曲线评估诊断性能。结果 224 例胸腰椎爆裂性骨折患者平均 LOS 为 (10.25±4.01) 天,64 例患者 (28.6%) 延长 LOS (> 12 天)。延长 LOS 患者年龄[(52.41±12.58) 岁 vs.(48.65±11.87) 岁]、BMI-偏瘦 (13.1% vs.7.2%)和肥胖 (15.1% vs.6.8%)、美国麻醉医师协会 (American Society of Anesthesiologists,ASA) 评分 (3~4 分)(25.0% vs.6.8%)、深静脉血栓形成 (deep vein thrombosis,DVT) (9.4% vs.2.5%)、甘油三酯 (triglyceride,TG)(> 1.7 mmol / L) (34.4% vs.19.4%)、白细胞 (white blood cell,WBC) (> 10×10~9 / L) (26.6% vs.11.3%)、C-反应蛋白 (C-reactive protein,CRP) (> 10 mg / L) (34.4% vs.17.5%)、血小板分布宽度 (platelet distribution width,PDW) (> 12%) (20.3% vs.7.5%)、淋巴细胞 (lymphocyte,LYM) (< 1.1×10~9 / L) (20.3% vs.10.0%) 和白蛋白(albumin,ALB) (< 35 g / L) (35.9% vs.20.6%) 比例高于常规 LOS (P < 0.05)。BMI-偏瘦和肥胖、ASA 评分(3~4 分)、糖尿病、CRP (> 10 mg / L) 和 ALB (< 35 g / L) 独立增加患者延迟 LOS 风险 (P < 0.05)。BMI、ASA 评分、糖尿病、CRP 和 ALB 的联合 ROC 的曲线下面积 (area under the cure,AUC) 为 0.842 (0.787~0.887)。结论BMI-偏瘦和肥胖、ASA 评分 (3~4 分)、糖尿病、CRP (> 10 mg / L) 和 ALB (< 35 g / L) 与胸腰椎爆裂性骨折患者延长 LOS 风险相关,且这些指标联合使用有较高预测延长 LOS 风险准确性。

Objective To analyze risk factors for prolonged length of stay (LOS) in patients with thoracolumbar burst fractures.Methods A retrospective analysis of the preoperative clinical data of 224 patients with thoracolumbar burst fractures who underwent long-segment endoprosthetic fixation between March 2019 and June 2023 at our hospital was applied.LOS greater than or equal to the 75th percentile was prolonged LOS.Independent risk factors for prolonged LOS in patients with thoracolumbar burst fractures were analyzed using thoracic and lumbar Logistic regression modelling.Diagnostic performance was assessed using combined receiver operating characteristic (ROC) curves.Results The mean LOS of 224 patients with thoracolumbar burst fractures was (10.25±4.01) d.64 patients (28.6%) had prolonged LOS (> 12 d).Age[(52.41±12.58) years vs.(48.65±11.87) years],BMI-lean(13.1% vs.7.2%) and obese (15.1% vs.6.8%),American Society of Anaesthesiologists (ASA) score (3-4)(25.0% vs.6.8%),deep vein thrombosis (DVT) (9.4% vs.2.5%),triglycerides (TG) (> 1.7 mmol / L) (34.4% vs.19.4%),white blood cells (WBC) (> 10 × 10~9 / L) (26.6% vs.11.3%),C-reactive protein (CRP) (> 10 mg / L)(34.4% vs.17.5%),platelet distribution width (PDW) (> 12%) (20.3% vs.7.5%),lymphocyte (LYM)(< 1.1 × 10~9 / L) (20.3% vs.10.0%),and ALB (< 35 g / L) (35.9% vs.20.6%) ratios of prolonged LOS were higher than those of the conventional LOS (P < 0.05).BMI-lean and obese,ASA score (3-4),diabetes,CRP (> 10 mg / L),CRP (> 10 mg / L),and ALB (< 35 g / L) were associated with the risk of prolonged LOS for patients with thoracolumbar burst fractures,and the combination of these metrics had a very high predictive accuracy.BMI-lean and obese,ASA score (3-4),diabetes mellitus,CRP (> 10 mg / L),and ALB (< 35 g / L) independently increased the risk of delayed LOS in patients (P < 0.05).The area under the cure (AUC) for the combined ROC of BMI,ASA score,diabetes mellitus,CRP,and ALB was 0.842 (0.787-0.887).Conclusions BMI-lean and obese,ASA score(3-4),diabetes mellitus,CRP (> 10 mg / L),CRP (> 10 mg / L),and ALB (< 35 g / L) are associated with the risk of prolonged LOS in patients with thoracolumbar burst fracture and the combination of these metrics indicates a high predictive accuracy for prolonged LOS risk.