骨质疏松性椎体压缩骨折患者术后疼痛缓解不佳风险列线图构建

Construction of nomogram of the risk of poor postoperative pain relief in patients with osteoporotic compression fractures

目的分析骨质疏松性椎体压缩骨折(osteoporotic vertebral compression fractures,OVCF)患者经皮椎体成形术(percutaneous vertebroplasty,PVP)术后疼痛缓解不佳的危险因素并构建列线图。方法选取2019年7月至2022年6月,我院PVP治疗的136例OVCF患者。根据术前1天和术后1个月的疼痛视觉模拟评分(visual analogue scales,VAS)差值定义疼痛缓解不佳和疼痛缓解良好。通过多因素Logistic回归分析OVCF患者术后疼痛缓解不佳的独立危险因素。通过接收者操作特征曲线(receiver operating characteristic curve,ROC)评估独立危险因素的诊断能力。基于独立危险因素列线图并通过校正曲线和决策曲线分析评估列线图预测准确性和临床净收益。结果PVP术后疼痛缓解不佳患者58例(42.6%),疼痛缓解良好患者78例(57.4%)。多因素Logistic回归分析结果显示,年龄、美国麻醉医师协会(American Society of Anesthesiologists,ASA)分级(Ⅲ~Ⅳ级)、吸烟、胸腰筋膜损伤、骨密度(bone mineral density,BMD)、骨水泥注入量和骨水泥渗漏是PVP术后疼痛缓解不佳的独立危险因素(P<0.05)。年龄、ASA分级、吸烟、胸腰筋膜损伤、BMD、骨水泥注入量和骨水泥渗漏曲线下面积(area under the curve,AUC)分别为0.782、0.604、0.601、0.622、0.743、0.840和0.651。PVP术后疼痛缓解不佳风险的列线图校正曲线与原始曲线及理想曲线接近,C-index为0.814(95%CI:0.704~0.913),模型拟合度高。当列线图预测风险阈值在0.08~1.00时,能提供显著意义的临床净收益。结论基于年龄、ASA分级、吸烟、胸腰筋膜损伤、BMD、骨水泥注入量和骨水泥渗漏构建的列线图能有效定量OVCF患者PVP术后疼痛缓解不佳风险。

Objective To analyse risk factors for poor pain relief after percutaneous vertebroplasty(PVP)in patients with osteoporotic vertebral compression fractures(OVCF)and to construct a column chart.Methods A total of 136 OVCF patients treated with PVP in our hospital from July 2019 to June 2022 were selected.Poor pain relief and good pain relief were defined according to the difference in visual analogue scales(VAS)of pain at 1 day preoperatively and 1 month postoperatively.Independent risk factors for poor postoperative pain relief in OVCF patients were analyzed by multifactorial logistic regression.Diagnostic ability of independent risk factors was assessed by receiver operating characteristic curve(ROC).Predictive accuracy and net clinical benefit of the column line diagrams were assessed based on independent risk factor column line diagrams and by calibration curve and decision curve analysis.Results There were 58 patients(42.6%)with poor pain relief and 78 patients(57.4%)with good pain relief after PVP.The results of multifactorial logistic regression analysis showed that age,ASA classification(gradeⅢ-Ⅳ),smoking,thoracolumbar fascial injury,bone mineral density(BMD)(T-score),bone cement injection volume,and cement leakage were the independent risk factors for poor pain relief after PVP(P<0.05).Age,ASA classification,smoking,thoracolumbar fascial injury,BMD(T-score),bone cement injection volume and bone cement leakage Area under the curve(AUC)were 0.782,0.604,0.601,0.622,0.743,0.840,and 0.651,respectively.Columnar plots of the risk of poor pain relief after PVP The calibration curve was close to the original and ideal curves,with a C-index of 0.814(95%CI:0.704 to 0.913)and a good model fit.The column-line plots provided a clinically significant net benefit when the predicted risk threshold was 0.08-1.00.Conclusions The nomogram constructed on the basis of age,ASA classification,smoking,thoracolumbar fascial injury,BMD(T-score),bone cement injection volume and cement leakage is effective in quantifying the risk of poor pain relief after PVP in patients with OVCF.