Development of a nomogram-based prediction model for fall risk after surgery for osteoporotic vertebral compression fractures

Objective To investigate the risk factors for increased fall risk following percutaneous vertebroplasty (PVP) in patients with osteoporotic vertebral compression fractures (OVCF) and to develop and validate a corresponding predictive nomogram. Methods In this prospective cohort study, a total of 297 patients diagnosed with OVCF who were scheduled to undergo PVP at our institution between June and October 2022 were consecutively enrolled. All enrolled patients were prospectively followed for one year postoperatively. A computer-generated random sequence was used to assign patients to a training cohort (n=208) or a validation cohort (n=89) in a 7: 3 ratio. Based on fall occurrences during follow-up, the training cohort was categorized into fall (n=83) and non-fall (n=125) groups. Univariate and multivariable binary logistic regression analyses identified independent risk factors, which were used to construct a nomogram. The model's discrimination, calibration, and clinical utility were assessed using the area under the receiver operating characteristic curve (AUC), calibration curves, and decision curve analysis (DCA), respectively. Results The one-year incidence of postoperative falls was 41. 41%(123/297). Multivariable analysis identified non-adherence to osteoporosis treatment (OR=3. 660, 95%CI: 1. 676-7. 991), history of falls within the past 6 months (OR=3. 056, 95%CI: 1. 505-6. 207), frailty (OR=4. 029, 95%CI: 1. 951-8. 321), sarcopenia (OR=2. 397, 95%CI: 1. 192-4. 819), non-use of walking aids (OR=2. 582, 95%CI: 1. 227-5. 434), impaired balance (OR=3. 055, 95%CI: 1. 438-6. 491), and nocturia (≥3 episodes/night) (OR=2. 723, 95%CI: 1. 293-5. 735) as independent risk factors (all P0. 05), and DCA showed a positive net clinical benefit. Conclusion Non-adherence to osteoporosis treatment, recent fall history, frailty, sarcopenia, non-use of walking aids, impaired balance, and frequent nocturia are independent risk factors for falls in OVCF patients after PVP. The developed nomogram provides a reliable tool for early identification of high-risk patients, facilitating targeted interventions. (目的 探讨骨质疏松性椎体压缩性骨折(OVCFs)患者行经皮椎体成形术(PVP)后跌倒风险增加的相关因素, 并构建及验证术后跌倒风险预测模型。方法 采用前瞻性队列研究设计, 于2022年6月—10月连续纳入中山市中医院骨四科收治的PVP手术患者297例。所有患者术后均接受为期1年的前瞻性随访。采用计算机生成随机序列, 以7: 3比例将患者分为建模集(n=208)与验证集(n=89)。依据随访期间是否发生跌倒, 将建模集患者进一步划分为跌倒组(n=83)与非跌倒组(n=125)。采用单因素及多因素二元Logistic回归分析筛选术后跌倒的独立危险因素, 据此构建列线图预测模型。通过受试者工作特征曲线下面积(AUC)、校准曲线及临床决策曲线(DCA)评估模型的区分度、校准能力与临床适用性。结果 PVP术后1年内, 患者跌倒发生率为41. 41%(123/297)。多因素分析显示, 未规律接受骨质疏松治疗(OR=3. 660, 95%CI: 1. 676-7. 991)、近6个月内有跌倒史(OR=3. 056, 95%CI: 1. 505-6. 207)、存在衰弱(OR=4. 029, 95%CI: 1. 951-8. 321)、存在肌少症(OR=2. 397, 95%CI: 1. 192-4. 819)、未使用助行工具(OR=2. 582, 95%CI: 1. 227-5. 434)、平衡能力不足(OR=3. 055, 95%CI: 1. 438-6. 491)及夜尿≥3次/晚(OR=2. 723, 95%CI: 1. 293-5. 735)为PVP术后跌倒的独立危险因素(均P＜0. 05)。基于上述因素构建的列线图模型在建模集(AUC=0. 824)与验证集(AUC=0. 849)中均表现出良好的预测效能。校准曲线与Hosmer-Lemeshow检验提示模型预测值与实际观测值一致性良好(P＞0. 05), DCA曲线显示模型具有积极的临床净收益。结论 未规律接受骨质疏松治疗、近6个月内跌倒史、衰弱、肌少症、未使用助行工具、平衡能力不足及夜尿频繁是OVCFs患者PVP术后跌倒的独立危险因素。本研究构建的列线图预测模型具有良好的区分度与准确度, 有助于临床早期识别高危患者并指导干预。)