Scholarlink

Objective <p>To systematically review published studies on risk prediction models for adjacent vertebral fractures (AVF) after vertebral augmentation (VA), thereby providing a reference for constructing and improving such models.</p> Methods <p>PubMed, Web of Science, The Cochrane Library, Embase, China National Knowledge Infrastructure (CNKI), China Science and Technology Journal Database (VIP), Wanfang Database, and SinoMed were searched from their inception to July 13, 2024. The Prediction Model Risk of Bias Assessment Tool (PROBAST) was used to assess the risk of bias and applicability of the prediction model studies; STATA 15.0 software was used to perform a meta-analysis on the area under the curve (AUC) values of the model validation and the common predictors used in model construction.</p> Results <p>A total of 13 studies were included, establishing 13 risk prediction models, with a total sample size of 3,083 patients. The AUC values of the included models ranged from 0.72 to 0.988. Of the included studies, 11 conducted internal validation, while two performed external validation. According to the PROBAST evaluation, all 13 studies exhibited a high risk of bias, yet demonstrated good applicability. The results of meta-analysis showed that the combined AUC value for the 5 validation models was 0.86 (95% CI: 0.76, 0.97). Notably, bone cement leakage (OR = 5.75, 95% CI: 3.43 ~ 9.60), age (OR = 1.20, 95% CI: 1.05 ~ 1.36), and a history of vertebral fractures (OR = 2.60, 95% CI: 1.64 ~ 4.13) were identified as significant high-risk factors for AVF after VA.</p> Conclusion <p>The risk prediction models for AVF after VA performed well, but exhibited a high risk of bias. It is recommended that future studies should consider selecting more appropriate machine learning algorithms and conducting large-sample, multicenter studies. Meanwhile, healthcare providers should focus on patients with bone cement leakage, advanced age, and a previous history of vertebral fractures, remaining vigilant for the potential occurrence of AVF.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Prediction models for adjacent vertebral fractures after vertebral augmentation: a systematic review and meta-analysis

  • 期刊论文

  • 发表日期：

    2025年3月17日

查看全文
  • Dan Sun,
  • Yuhang Wen,
  • Qiongge Yu,
  • Yu Long,
  • Yuyan Liu,
  • Yue Zhou,
  • Yufeng Yu

摘要

Objective

To systematically review published studies on risk prediction models for adjacent vertebral fractures (AVF) after vertebral augmentation (VA), thereby providing a reference for constructing and improving such models.

Methods

PubMed, Web of Science, The Cochrane Library, Embase, China National Knowledge Infrastructure (CNKI), China Science and Technology Journal Database (VIP), Wanfang Database, and SinoMed were searched from their inception to July 13, 2024. The Prediction Model Risk of Bias Assessment Tool (PROBAST) was used to assess the risk of bias and applicability of the prediction model studies; STATA 15.0 software was used to perform a meta-analysis on the area under the curve (AUC) values of the model validation and the common predictors used in model construction.

Results

A total of 13 studies were included, establishing 13 risk prediction models, with a total sample size of 3,083 patients. The AUC values of the included models ranged from 0.72 to 0.988. Of the included studies, 11 conducted internal validation, while two performed external validation. According to the PROBAST evaluation, all 13 studies exhibited a high risk of bias, yet demonstrated good applicability. The results of meta-analysis showed that the combined AUC value for the 5 validation models was 0.86 (95% CI: 0.76, 0.97). Notably, bone cement leakage (OR = 5.75, 95% CI: 3.43 ~ 9.60), age (OR = 1.20, 95% CI: 1.05 ~ 1.36), and a history of vertebral fractures (OR = 2.60, 95% CI: 1.64 ~ 4.13) were identified as significant high-risk factors for AVF after VA.

Conclusion

The risk prediction models for AVF after VA performed well, but exhibited a high risk of bias. It is recommended that future studies should consider selecting more appropriate machine learning algorithms and conducting large-sample, multicenter studies. Meanwhile, healthcare providers should focus on patients with bone cement leakage, advanced age, and a previous history of vertebral fractures, remaining vigilant for the potential occurrence of AVF.