目的 系统评价重症监护室（intensive care unit，ICU）患者静脉血栓栓塞症风险预测模型，为临床护理及未来研究提供参考。方法 检索PubMed、Web of Science、Cochrane library、Wiley、CINAHL、SinoMed、中国知网、万方中发表的ICU患者静脉血栓栓塞症风险预测模型研究，检索时限为建库至2025年3月31日。由2名研究者独立筛选文献、提取数据,并对纳入文献进行偏倚风险评价、适用性评价。结果 共纳入21项研究，其中曲线下面积＞0.8的模型占66.67%，模型总体预测性能良好，但整体偏倚风险高，少数模型进行外部验证。出现频次最高的3项预测因子为既往静脉血栓栓塞症史、D-二聚体及制动。结论 目前ICU患者静脉血栓栓塞症风险预测模型仍有很大的发展空间；未来可参考PROBAST指南，建立偏倚风险低、预测性能好、决策透明度高、临床适用性好的模型并进行广泛的外部验证，同时推动模型应用与临床实践。

Objective To systematically review risk prediction models for venous thromboembolism (VTE) in ICU patients，providing reference for clinical nursing practice and future research.Methods PubMed，Web of Science，Cochrane library，Wiley，CINAHL，SinoMed，CNKI，and Wanfang Data were searched for relative studies from inception to March 31，2025.2 researchers independently screened the literature，extracted data，and assessed the risk of bias and applicability of the included studies.Results 21 studies were included.Among these，66.67% of the models had an area under the curve greater than 0.8，indicating generally good overall predictive performance.However，the overall risk of bias was high，and only a minority of models underwent external validation.The 3 most frequently occurring predictors were history of VTE，D-dimer levels，and immobilization.Conclusions Current risk prediction models for VTE in ICU patients are still open to development.Future research should refer to the PROBAST guidelines to develop models with low risk of bias，good predictive performance，high decision transparency，and strong clinical applicability，accompanied by extensive external validation.Simultaneously，efforts should be made to promote the application and integration of these models into clinical practice.

R47；R473.54