首页|桡骨远端骨折切开复位掌侧钢板内固定术后骨折愈合时间分布及延迟愈合的因素探讨

桡骨远端骨折切开复位掌侧钢板内固定术后骨折愈合时间分布及延迟愈合的因素探讨

扫码查看
目的 探究影响桡骨远端骨折掌侧钢板切开复位内固定术治疗术后骨折愈合时间分布及延迟愈合的因素.方法 选取 2019 年 6 月至 2022 年 4 月,在我院进行治疗的 94 例桡骨远端骨折的患者为研究对象,进行掌侧钢板切开复位内固定术治疗.记录患者的基线资料,根据桡骨远端骨折患者术后愈合时间长短将其分为愈合正常组和愈合延迟组.采用 LASSO 回归分析筛选影响桡骨远端骨折术后愈合时间的因素,并采用 Logistic 回归分析模型和决策树模型进行分析.采用多重共线性诊断分析 Logistic 回归模型多因素间的共线性,并利用受试者工作特征(ROC)曲线对两个模型的分析结果进行比较.构建列线图模型并评价其准确性和临床实用性.结果 本组 94 例中,有 77.66%(73/94)的患者术后正常愈合,愈合时间为(2.6±0.3)个月;22.34%(21/94)的患者术后愈合时间延长,愈合时间为(3.9±0.5)个月.LASSO 回归分析筛选出 6 个潜在的危险因素,分别为患者年龄、骨折类型、骨质疏松情况、桡骨重度缩短畸形情况、功能锻炼情况以及医源性软组织损伤情况.Logistic 回归分析结果显示,患者的年龄、骨折类型、骨质疏松情况、桡骨重度缩短畸形情况以及功能锻炼情况是影响桡骨远端骨折患者术后愈合时间是否延长的危险因素,并且这些因素的方差膨胀因子(VIF)均<5,各因素之间不存在共线性.决策树模型分析显示骨折类型、患者的年龄、骨质疏松情况以及功能锻炼情况是影响桡骨远端骨折患者术后愈合时间的影响因素.两模型分析结果显示,Logistic 回归分析模型的灵敏度为 79.6%,特异度为 82.4%;决策树模型的灵敏度 77.5%,特异度为 87.6%.Logistic 回归分析模型的 AUC 为 0.838(95%CI:0.814~0.883,P<0.001),决策树模型的 AUC 为 0.829(95%CI:0.802~0.861,P<0.001),两个模型均呈现良好的预测性.列线图模型的准确性和临床实用性均较高.结论 患者年龄、骨折类型、骨质疏松情况、桡骨重度缩短畸形情况以及功能锻炼情况是影响桡骨远端骨折患者掌侧钢板切开复位内固定术治疗术后愈合时间长短的危险因素.
Investigation of fracture healing time distribution and delayed healing factors after volar plate open reduction and internal fixation for distal radius fractures
Objective To explore the factors influencing the fracture healing time distribution and delayed healing after volar plate open reduction and internal fixation for distal radius fractures.Methods A series of 94 patients with distal radius fracture who received treatment in our hospital from June 2019 to April 2022 were selected as research objects and treated with volar plate open reduction and internal fixation.The baseline data of patients were recorded.According to the postoperative healing time of patients with distal radius fracture,all were divided into normal healing group and delayed healing group.LASSO regression analysis was used to screen the factors affecting the postoperative healing time of distal radius fractures.Logistic regression analysis model and decision tree model were used to analyze the factors.The collinearity among multiple factors of Logistic regression model was analyzed using multicollinearity diagnosis,and the analysis results of the two models were compared by receiver operating characteristic(ROC)curve.Construct the line chart model and evaluate its accuracy and clinical practicability.Results Among 94 patients with distal radius fracture,77.66%(73/94)of them healed normally after surgery,with a healing time of(2.6±0.3)months.22.34%(21/94)of them had prolonged postoperative healing time,with healing time of(3.9±0.5)months.Six potential risk factors were identified by LASSO regression analysis:age,fracture type,osteoporosis,severe radius shortening deformity,functional exercise,and iatrogenic soft tissue injury.Logistic regression analysis showed that the age,fracture type,osteoporosis,severely shortened radius deformity and functional exercise were risk factors for prolonged postoperative healing time in patients with distal radius fracture,and the variance inflation factor(VIF)of these factors were all less than 5,and there was no collinearity among all factors.The analysis of decision tree model showed that fracture type,patient age,osteoporosis and functional exercise were influencing factors on postoperative healing time of patients with distal radius fracture.The sensitivity and specificity of Logistic regression analysis model were 79.6%and 82.4%respectively.The sensitivity and specificity of decision tree model were 77.5%and 87.6%respectively.The AUC of Logistic regression analysis model was 0.838(95%CI:0.814-0.883,P<0.001),the AUC of the decision tree model was 0.829(95%CI:0.814-0.883,P<0.001).Both models showed good predictability.The accuracy and clinical practicability of the line chart model were high.Conclusions Patient age,fracture type,osteoporosis,severely shortened radius deformity,and functional exercise are risk factors for postoperative healing time in patients with distal radius fracture treated with volar plate open reduction and internal fixation.

Radius fractureFracture healingRisk factorsDecision treesNomogramsModels,statistical

郭洋、刘凯

展开 >

102600 北京市大兴区人民医院创伤科

桡骨骨折 骨折愈合 危险因素 决策树 列线图 模型,统计学

2024

中国骨与关节杂志
中国医疗保健国际交流促进会,北京中科康辰骨关节伤病研究所

中国骨与关节杂志

CSTPCD
影响因子:0.665
ISSN:2095-252X
年,卷(期):2024.13(8)
  • 3