| 吴维,黄杨,王晓铃,李佳.构建并验证预测骨质疏松性椎体压缩骨折骨水泥手术术后功能的列线图模型[J].中国康复,2023,38(4):226-231 |
| 构建并验证预测骨质疏松性椎体压缩骨折骨水泥手术术后功能的列线图模型 |
| Construction and validation of a nomographic model for predicting postoperative function recovery in patients with osteoporotic vertebral compression fractures |
| |
| DOI： |
| 中文关键词: 骨质疏松性椎体压缩骨折 Oswestry功能障碍评分 危险因素 列线图预测模型 |
| 英文关键词: osteoporotic vertebral compression fracture Oswestry disability index risk factors nomogram prediction model |
| 基金项目: |
|
| 摘要点击次数: 3528 |
| 全文下载次数: 8237 |
| 中文摘要: |
| 目的：基于Logistic回归分析得出影响骨质疏松性椎体压缩骨折患者行微创骨水泥手术术后功能的危险因素，并构建列线图预测模型。方法：回顾性分析2017年1月~2020年6月在我院施行微创骨水泥手术的骨质疏松性椎体压缩骨折(OVCF)患者共472例，根据术后6个月的Oswestry功能障碍评分（ODI）分为恢复良好组和恢复不良组。患者按7：3的比例被随机分为训练集（n=330）与测试集（n=142）。首先在训练集中，基于单因素分析及多因素Logistic回归分析筛选影响OVCF患者术后功能的危险因素，并构建列线图预测模型。随后，分别在训练集和测试集中对列线图的预测结果进行交叉验证，采用ROC曲线、校准曲线验证该模型的区分度及一致性，采用决策曲线验证对临床决策的净收益。最后，使用Pearson相关系数衡量列线图评分与临床实际参数的相关性。结果：年龄（> 67岁）、骨密度（≤ -2.5）、骨水泥注入量（> 5.0 mL）、椎体高度恢复率（≤ 38.45%）、Cobb角（> 15.08°）、碱性磷酸酶（≤ 82.69 IU/L）、血清骨代谢标志物-Ⅰ型前胶原氨基端前肽P1NP（≤ 34.57 μg/L）及术后不接受康复训练是影响OVCF患者椎体功能恢复的危险因素（均P＜0.05）。构建的列线图预测模型具有较好的区分度（ROC曲线下面积在训练集和测试集中分别为0.869和0.843）和一致性（Hosmer-Lemeshow拟合优度检验=5.918，P=0.656），且具有较高的净收益。相关性分析提示该列线图评分与Cobb角、年龄、骨水泥注入量正相关，与BMD、椎体高度恢复率、ALP、P1NP及接受康复训练呈负相关。结论：本研究构建的列线图模型具有较好的准确度与稳健性，可以帮助临床预测OVCF患者术后功能，提前干预高危患者。 |
| 英文摘要: |
| Objective: To investigate the risk factors affecting postoperative function of patients with osteoporotic vertebral compression fracture (OVCF) and to construct a nomogram prediction model based on Logistic regression analysis. Methods: Totally, 472 OVCF patients who underwent bone cement surgery in our hospital from January 2017 to June 2020 were retrospectively analyzed. According to the Oswestry disability index (ODI) at 6th month after surgery, they were divided into good recovery group and poor recovery group. Patients were randomly divided into an independent training set (n=330) and a test set (n=142) in a ratio of 7:3. In the independent training set, single-factor and multi-factor logistic regression analyses were used to screen the risk factors affecting poor recovery of vertebral function in OVCF patients, and a nomogram prediction model was constructed. Subsequently, the nomogram was cross-validated in the independent training set and the test set, respectively, using receiver operating characteristics (ROC) curves and calibration curves to verify the consistency and discrimination, and decision curves to verify the net benefit to clinical decision making. Finally, Pearson correlation coefficients were used to measure the correlation between nomogram scores and clinical factors. Results: Patients in the poor recovery group had significantly higher rate of smoking history, less cement injection volume, higher BMD, higher proportion of vertebral scoliosis deformity, and older age, greater Cobb angle, and higher VAS score than those in the good recovery group (P<0.05), and the preoperative height of the injured spine, recovery rate of vertebral height, ALP, and P1NP were significantly lower than those in the good recovery group(P<0.05). The results of multifactorial logistic re-gression analysis showed that cement injection volume (> 4.50 mL), BMD (≤-3.50 g/m3), age (> 67 years), vertebral height recovery rate (≤38.45%), Cobb angle (>15.08°), ALP (≤82.69 IU/L), P1NP (≤ 34.57 μg/L) and without receiving rehabilitation training after surgery were the most important factors affecting risk factors for poor recovery of vertebral function in OVCF patients (P<0.05). The constructed nomogram prediction model had good discrimination (area under the ROC curve was 0.869 in the training set and 0.843 in the test set) and consistency (Hosmer-Lemeshow goodness-of-fit test=5.918, P=0.656) with a high net benefit. Correlation analysis suggested that the risk of poor vertebral function recovery in OVCF patients predicted by nomogram was positively correlated with Cobb angle, age, and cement injection, and negatively correlated with BMD, vertebral height recovery rate, ALP, P1NP and rehabilitation training given after surgery. Conclusion: The nomogram model constructed in this study to predict the functional recovery in OVCF patients has good accuracy and robustness and can be used as an aid to clinical decision making. |
|
查看全文
下载PDF阅读器 HTML全文 |
| 关闭 |
|
|
|
