Zhou Lin-Feng, Jiang Tao, Zhang Xiao-Qin, Li Zhi-Rong
Orthopedic Surgery Department 2, Ya'an Hospital of Traditional Chinese Medicine, Ya' an, Sichuan, China.
Medicine (Baltimore). 2025 Jun 27;104(26):e43096. doi: 10.1097/MD.0000000000043096.
This study was intended to analyze the potential risk factors associated with postoperative onset of kinesiophobia in patients with lumbar disc herniation and to construct a predictive model using logistic regression and decision tree methods. We retrospectively evaluated the clinical data of 342 lumbar disc herniation patients who underwent surgical treatment between January 2021 and December 2023, grouped them according to the occurrence of postoperative kinesiophobia, applied multifactorial logistic regression to determine the main influencing factors, and constructed a prediction model with the help of SPSS Modeler software to further evaluate its predictive effect. The study found that the incidence of kinesiophobia was 37.46%; statistically significant differences were found between the kinesiophobia group and the nonkinesiophobia group in terms of education level, visual analog scale (VAS) score of pain, Hospital Anxiety and Depression Scale (HADS) score, self-efficacy, per capita monthly family income, and mode of payment of medical expenses (P < .05). Multifactorial Logistic regression suggested that all of the above factors were independent influencing variables of kinesiophobia (P < .05). Decision tree modeling revealed that self-efficacy was the first discriminant variable, followed by payment method, VAS, HADS score, and income level. Analysis of the subjects' job characteristic curve (receiver operating characteristic) showed that the predictive power of the decision tree model was significantly better than that of the logistic regression model (P < .05). Education level, VAS score, HADS score, self-efficacy, family financial status, and payment mode are all important risk factors for postoperative kinesiophobia, and the use of predictive modeling can be more effective in assessing the patient's condition and realizing early intervention.
本研究旨在分析腰椎间盘突出症患者术后发生运动恐惧的潜在危险因素，并运用逻辑回归和决策树方法构建预测模型。我们回顾性评估了2021年1月至2023年12月期间接受手术治疗的342例腰椎间盘突出症患者的临床资料，根据术后是否发生运动恐惧进行分组，应用多因素逻辑回归确定主要影响因素，并借助SPSS Modeler软件构建预测模型，以进一步评估其预测效果。研究发现，运动恐惧的发生率为37.46%；运动恐惧组与非运动恐惧组在教育水平、视觉模拟评分法（VAS）疼痛评分、医院焦虑抑郁量表（HADS）评分、自我效能感、家庭人均月收入以及医疗费用支付方式方面存在统计学显著差异（P < .05）。多因素逻辑回归表明，上述所有因素均为运动恐惧的独立影响变量（P < .05）。决策树建模显示，自我效能感是首要判别变量，其次是支付方式、VAS评分、HADS评分和收入水平。对受试者工作特征曲线（ROC曲线）的分析表明，决策树模型的预测能力显著优于逻辑回归模型（P < .05）。教育水平、VAS评分、HADS评分、自我效能感、家庭经济状况和支付方式均是术后运动恐惧的重要危险因素，运用预测模型可更有效地评估患者病情并实现早期干预。
