口腔医学研究

口腔医学研究 ›› 2024, Vol. 40 ›› Issue (10): 890-894.DOI: 10.13701/j.cnki.kqyxyj.2024.10.008

• 口腔颌面外科学研究 • 上一篇    下一篇

颌面部骨折术后感染的危险因素分析及风险模型构建

陆欣悦1, 潘玥彤1, 孙昕奕1, 吕中静1,2*   

  1. 1.徐州医科大学口腔医学院 江苏 徐州 221004;
    2.徐州医科大学附属医院口腔科 江苏 徐州 221006
  • 收稿日期:2024-05-09 出版日期:2024-10-28 发布日期:2024-10-24
  • 通讯作者: *吕中静,E-mail:Zhongjing_lv2012@163.com
  • 作者简介:陆欣悦(2003~),女,江苏常州人,本科,研究方向:口腔颌面部肿瘤、口腔颌面部骨折。
  • 基金资助:
    徐州市科技计划基金项目(编号:KC22217)

Risk Factor Analysis and Risk Model Construction of Postoperative Infection in Patients with Maxillofacial Fractures

LU Xinyue1, PAN Yuetong1, SUN Xinyi1, LV Zhongjin1,2*   

  1. 1. School of Stomatology, Xuzhou Medical University, Xuzhou 221004, China;
    2. Department of Stomatology, The Affiliated Xuzhou Municipal Hospital of Xuzhou Medical University, Xuzhou 221006, China
  • Received:2024-05-09 Online:2024-10-28 Published:2024-10-24

PDF (PC)

10

摘要： 目的: 分析颌面部骨折患者术后感染的危险因素,并构建风险预测模型。方法: 选取2022年1月~2024年1月徐州医科大学附属医院收治的口腔颌面部骨折行手术治疗后发生感染的81例患者作为研究对象,另选取同期于该院收治的口腔颌面部骨折行手术治疗后未发生感染的70例患者作为对照组,收集患者临床资料并进行多因素Logistic回归分析筛选独立危险因素,构建风险预测模型并评估其预测价值。结果: 研究发现手术时长≥3 h、合并糖尿病、钛钉使用数量≥20个是颌面部骨折术后感染的独立危险因素(P<0.05);依据筛选的独立风险因素构建颌面部骨折患者术后感染风险预测模型:Logit(P)=-0.747+手术时长×1.730+合并糖尿病×1.789+钛钉使用数量×1.078,Hosmer-Lemeshow拟合优度检验提示模型的拟合效果较好(χ2=2.015,P=0.365),校准曲线显示预测概率与实际概率接近,提示模型具有较好的校准度;受试者工作特征曲线显示预测模型的曲线下面积值为0.728,提示模型具有一定区分能力;决策曲线分析显示在0.3~0.8的横坐标范围内,预测模型曲线位于两条极端曲线的上方,提示模型的临床实用性尚可。结论: 颌面部骨折患者术后感染与手术时长、合并糖尿病、钛钉使用数量有关,据此构建的风险预测模型具有一定的诊断价值。

关键词: 颌面部骨折, 术后感染, 危险因素, 预测模型, 诊断价值

Abstract: Objective: To analyze the risk factors for postoperative infection in patients with maxillofacial fractures and construct a risk prediction model. Methods: From January 2022 to January 2024, 81 patients with oral and maxillofacial fractures who developed infection after surgical treatment at Xuzhou Medical University Affiliated Hospital were selected as the research subjects. Meanwhile, 70 patients with oral and maxillofacial fractures who did not develop infection after surgery at the same hospital during the same period were selected as the control group. Clinical data of patients were retrospectively collected and multiple logistic regression analysis was performed to screen for independent risk factors. A risk prediction model was constructed and its predictive value was evaluated. Results: The study found that operation duration ≥3 h, diabetes, and the number of titanium nails ≥20 were independent risk factors for postoperative infection of maxillofacial fractures (P<0.05). According to the selected independent risk factors, a prediction model for postoperative infection risk of patients with maxillofacial fractures was constructed: Logit (P) =-0.747+operation duration×1.730+diabetes×1.789+number of titanium nails×1.078. Hosmer-Lemeshow good of fit test showed that the fitting accuracy of the model was good (χ2=2.015, P=0.365). The calibration curve showed that the prediction probability was close to the actual probability, indicating that the model had a good calibration degree. The receiver operating characteristic (ROC) curve indicated that the area under the curve (AUC) of the predictive model was 0.728, suggesting that the model possessed moderate discriminative capacity. Decision curve analysis (DCA) revealed that within the horizontal axis range of 0.3-0.8 the predictive model's curve lay above the two extreme curves, suggesting that the model's clinical utility was acceptable. Conclusion: Postoperative infection in patients with maxillofacial fractures is related to the operation duration, diabetes, and the number of titanium nails. The risk prediction model based on it has certain diagnostic value.

Key words: maxillofacial fractures, postoperative infection, risk factors, prediction model, diagnostic value