Objective  To systematically review the bleeding risk prediction models for patients with non-valvular atrial fibrillation (NVAF) taking new oral anticoagulants, and to provide references for constructing precise and practical prediction models.

Methods  PubMed, Cochrane Library, Scopus, Web of Science, EBSCO, Embase, CNKI, WanFang Data, VIP and SinoMed databases were electronically searched to collect studies on the construction or validation of bleeding risk prediction models for patients with NVAF after taking new oral anticoagulants from inception to July 16, 2025. Two researchers independently screened the literature, extracted data in accordance with the Critical Appraisal and Data Extraction for Systematic Reviews of Prediction Modelling Studies (CHARMS), and assessed the risk of bias and applicability using the Prediction Model Risk of Bias Assessment Tool (PROBAST).

Results  A total of 13 studies were included, with a sample size ranging from 237 to 55,903 cases, and outcome events ranging from 18 to 2,238 cases. Among these cases, 12 studies reported the discrimination, and 5 studies reported the calibration, only 2 studies performed both internal and external validation. In terms of overall risk of bias, all 13 studies were assessed as high risk. Regarding concerns about applicability, 7 studies showed low risk, 4 studies showed high risk, and 2 studies had unclear risk.

Conclusion  There are still many deficiencies in the existing prediction models of bleeding risk in patients with NVAF who are taking new oral anticoagulants. The main issues focus on the study population, predictors, and statistical analysis methodology. Future efforts should focus on developing models in accordance with guidelines; employing prospective cohort designs with sufficient follow-up to include an adequate number of outcome events; and implementing diverse modeling approaches and presentation formats. This will enable the construction of predictive models with better performance, so as to provide a more reliable support for clinical decision-making.

1.Brod C, Groth N, Rudeck M, et al. Oral anticoagulation use in non-valvular atrial fibrillation patients in rural setting[J]. Am J Med Open, 2023, 9: 100026. DOI: 10.1016/j.ajmo.2022.100026.

2.Ma C, Wu S, Liu S, et al. Chinese guidelines for the diagnosis and management of atrial fibrillation[J]. Pacing Clin Electrophysiol, 2024, 47(6): 714-770. DOI: 10.1111/pace.14920.

3.张献文, 郑燎源, 经煜, 等. 非瓣膜性房颤所致轻中型卒中复发危险因素分析[J]. 四川医学, 2024, 45(3): 225-229.[Zhang XW, Zheng LY, Jing Y, et al. Analysis of risk factors for recurrence of mild and moderate stroke caused by non-valvular atrial fibrillation[J]. Sichuan Medical Journal, 2024, 45(3): 225-229.] DOI: 10.16252/j.cnki.issn1004-0501-2024.03.001.

4.Liu X, Feng G, Marler SV, et al. Real world time trends in antithrombotic treatment for newly diagnosed atrial fibrillation in China: reports from the GLORIA-AF Phase Ⅲ registry: trends in antithrombotic therapy use in China[J]. Thromb J, 2023, 21(1): 83. DOI: 10.1186/s12959-023-00527-x.

5.Rao S, Aggarwal S, Mani S, et al. Uncovering the role of direct oral anticoagulants in stroke prevention for atrial fibrillation: a review of the literature[J]. Cureus, 2024, 16(7): e63675. DOI: 10.7759/cureus.63675.

6.Kato Y, Tsutsui K, Nakano S, et al. Cardioembolic stroke: past advancements, current challenges, and future directions[J]. Int J Mol Sci, 2024, 25(11): 5777. DOI: 10.3390/ijms25115777.

7.Campello E, Spiezia L, Simion C, et al. Direct oral anticoagulants in patients with inherited thrombophilia and venous thromboembolism: a prospective cohort study[J]. J Am Heart Assoc, 2020, 9(23): e018917. DOI: 10.1161/JAHA.120.018917.

8.Lyu M, Jiang S, Wu T, et al. A new model to predict the risk of major gastrointestinal bleeding in patients on direct oral anticoagulants (dabigatran and rivaroxaban)[J]. Br J Cli Pharmacol, 2023, 89(1): 253-260. DOI: 10.1111/bcp.15491.

9.Gao X, Cai X, Yang Y, et al. Diagnostic accuracy of the HAS-BLED bleeding score in VKA- or DOAC-treated patients with atrial fibrillation: a systematic review and Meta-analysis[J]. Front Cardiovasc Med, 2021, 8: 757087. DOI: 10.3389/fcvm.2021. 757087.

10.Moons KGM, De Groot JAH, Bouwmeester W, et al. Critical appraisal and data extraction for systematic reviews of prediction modelling studies: the CHARMS checklist[J]. PLoS Med, 2014, 11(10): e1001744. DOI: 10.1371/journal.pmed.1001744.

11.Moons KGM, Wolff RF, Riley RD, et al. PROBAST: a tool to assess risk of bias and applicability of prediction model studies: explanation and elaboration[J]. Annal Intern Med, 2019, 170(1): W1-W33. DOI: 10.7326/M18-1377.

12.Rutherford OCW, Jonasson C, Ghanima W, et al. New score for assessing bleeding risk in patients with atrial fibrillation treated with NOACs[J]. Open Heart, 2018, 5(2): e000931. DOI: 10.1136/openhrt-2018-000931.

13.Huang YC, Cheng YC, Jhou MJ, et al. Important risk factors in patients with nonvalvular atrial fibrillation taking dabigatran using integrated machine learning scheme-a post hoc analysis[J]. J Pers Med, 2022, 12(5): 756. DOI: 10.3390/jpm12050756.

14.Chen J, Lyu M, Xu W, et al. New score for predicting major bleeding in patients with atrial fibrillation using direct oral anticoagulants[J]. Inter J Cardiol, 2023, 376: 56-61. DOI: 10.1016/j.ijcard.2023.02.017.

15.Cao C, Xu Y, Jiang W, et al. Nomogram for predicting bleeding events in nonvalvular atrial fibrillation patients receiving rivaroxaban: a retrospective study[J]. Health Sci Rep, 2024, 7(1): e1792. DOI: 10.1002/hsr2.1792.

16.Barnett-Griness O, Stein N, Kotler A, et al. Novel bleeding prediction model in atrial fibrillation patients on new oral anticoagulants[J]. Heart, 2022, 108(4): 266-273. DOI: 10.1136/heartjnl-2021-319702.

17.AlAmmari M, Sultana K, Alturaiki A, et al. The development and validation of a multivariable model to predict the bleeding risk score for patients with non-valvular atrial fibrillation using direct oral anticoagulants in the arab population[J]. PLoS One, 2021, 16(5): e0250502. DOI: 10.1371/journal.pone.0250502.

18.Aggarwal R, Ruff CT, Virdone S, et al. Development and validation of the DOAC score: a novel bleeding risk prediction tool for patients with atrial fibrillation on direct-acting oral anticoagulants[J]. Circulation, 2023, 148(12): 936-946. DOI: 10.1161/CIRCULATIONAHA.123.064556.

19.Chaudhary R, Nourelahi M, Thoma FW, et al. Machine learning predicts bleeding risk in atrial fibrillation patients on direct oral anticoagulant[J]. Am J Cardiol, 2025, 244: 58-66. DOI: 10.1016/j.amjcard.2025.02.030.

20.朱子云, 张莲红, 刘宇恒. 非瓣膜性房颤患者服用达比加群酯后出血风险预测模型构建[J]. 药品评价, 2024, 21(3): 348-352.[Zhu ZY, Zhang LH, Liu YH. Construction of bleeding risk prediction model for patients with non-valvular atrial fibrillation after taking dabigatran etexilate[J]. Drug Eval, 2024, 21(3): 348-352.] DOI: 10.19939/j.cnki.1672-2809.2024.03.23.

21.覃润. 非瓣膜性房颤患者使用利伐沙班发生出血的危险因素分析[D]. 贵阳: 贵州医科大学, 2023. DOI: 10.27045/d.cnki.ggyyc.2023.000484.

22.葛友婷, 宋伟华, 高岸, 等. 非瓣膜性房颤患者服用新型口服抗凝药致消化道出血的影响因素分析及预测模型建立[J]. 上海医药, 2023, 44(19): 79-82, 87. [Ge YT, Song WH, Gao A, et al. Analysis of influencing factors and establishment of prediction model for gastrointestinal bleeding caused by new oral anticoagulants in patients with nonvalvular atrial fibrillation[J]. Shanghai Medical & Pharmaceutical Journal, 2023, 44(19): 79-82, 87.] DOI: 10.3969/j.issn.1006-1533.2023.19.016.

23.仲璟, 王泽旭, 王宏军, 等. 基于CES1基因多态性建立非瓣膜性房颤患者达比加群酯相关消化道出血风险模型 [J]. 中国老年学杂志, 2025, 45(5): 1046-1049. DOI: 10.3969/j.issn.1005-9202.2025.05.007.

24.张晨, 席小青, 高银凤, 等. 基于基因多态性构建的老年心房颤动患者利伐沙班相关出血事件风险模型分析[J]. 中华老年心脑血管病杂志, 2025, 27(1): 13-17. [Zhang C, Xi XQ, Gao YF, et al. Analysis of a risk model for rivaroxaban-related bleeding events in elderly patients with atrial fibrillation based on gene polymorphism[J]. Chinese Journal of Geriatric Heart Brain and Vessel Diseases, 2025, 27(1): 13-17.] DOI: 10.3969/ji.ssn.1009-0126.2025.01.004.

25.鲁小丹, 卫建华, 沈建通, 等. 预测模型系统评价的制作方法与步骤[J]. 中国循证医学杂志, 2023, 23(5): 602-609. [Lu XD, Wei JH, Shen JT, et al. Methods and processes for producing a systematic review of predictive model studies[J]. Chinese Journal of Evidence-Based Medicine, 2023, 23(5): 602-609.] DOI: 10.7507/1672-2531.202212112.

26.Efthimiou O, Seo M, Chalkou K, et al. Developing clinical prediction models: a step-by-step guide[J]. BMJ, 2024: e078276. DOI: 10.1136/bmj-2023-078276.

27.熊中敏, 郭怀宇, 吴月欣. 缺失数据处理方法研究综述[J]. 计算机工程与应用, 2021, 57(14): 27-38. [Xiong ZM, Guo HY, Wu XY. Review of missing data processing methods[J]. Comput Eng Appl, 2021, 57(14): 27-38.] DOI: 10.3778/j.issn.1002-8331. 2101-0187.

28.王胜锋, 王俊峰, 谷鸿秋, 等主编. 临床预测模型方法与应用[M]. 北京: 北京大学医学出版社, 2024: 2-450.

29.吴艳, 艾芬. 原发性高血压患者并发左心室肥厚的预测模型构建与验证[J]. 数理医药学杂志, 2024, 37(3): 190-200. [Wu Y, Ai F. Construction and validation of a prediction model for left ventricular hypertrophy in patients with essential hypertension[J]. Journal of Mathematical Medicine, 2024, 37(3): 190-200.] DOI: 10.19960/j.issn.1004-4337.202312067.

30.Ren J, Wang N, Zhang X, et al. A systematic review and meta-analysis of the morbidity of efficacy endpoints and bleeding events in elderly and young patients treated with the same dose rivaroxaban[J]. Ann Hematol, 2024, 103(11): 4363-4373. DOI: 10.1007/s00277-024-05767-z.

31.Wu VCC, Wang CL, Huang YC, et al. Bleeding associated with antiarrhythmic drugs in patients with atrial fibrillation using direct oral anticoagulants: a nationwide population cohort study[J]. J Am Heart Assoc, 2024, 13(21): e033513. DOI: 10.1161/JAHA.123. 033513.

32.别自东, 许莉莉, 赵庆恒, 等. 山东省不同区域医院华法林引起出血的临床特征及影响因素[J]. 中华高血压杂志（中英文）, 2024, 32(10): 985-991. DOI: 10.16439/j.issn.1673-7245.2024. 10.014.

33.Zhou Q, Liu X, Yang X, et al. Efficacy and safety of anticoagulation in atrial fibrillation patients with intracranial hemorrhage: a systematic review and Meta-analysis[J]. Front Pharmacol, 2023, 14: 1122564. DOI: 10.3389/fphar.2023.1122564.

34.Spruit JR, de Vries TAC, Hemels MEW, et al. Direct oral anticoagulants in older and frail patients with atrial fibrillation: a decade of experience[J]. Drugs Aging, 2024, 41(9): 725-740. DOI: 10.1007/s40266-024-01138-5.

35.Kikuchi T, Kono Y, Nakagawa K, et al. Clinical significance of gastrointestinal bleeding history in patients who undergo left atrial appendage closure[J]. JGH Open, 2024, 8(1): e13009. DOI: 10.1002/jgh3.13009.

36.Garcia DA, Fisher DA, Mulder H, et al. Gastrointestinal bleeding in patients with atrial fibrillation treated with apixaban or warfarin: insights from the apixaban for reduction in stroke and other thromboembolic events in atrial fibrillation (ARISTOTLE) trial[J]. Am Heart J, 2020, 221: 1-8. DOI: 10.1016/j.ahj.2019.10.013.

37.章少颖, 盛吉莅. 应用抗凝及抗血小板药物后针灸治疗的风险与对策[J]. 针刺研究, 2021, 46(8): 707-709, 716. [Zhang SY, Sheng JL. Hemorrhage risk and counter measures of acupuncture treatment in patients undergoing administration of anticoagulants and antiplatelets[J]. Acupuncture Research, 2021, 46(8): 707-709, 716.] DOI: 10.13702/j.1000-0607.20210151.

38.Ramspek CL, Jager KJ, Dekker FW, et al. External validation of prognostic models: what, why, how, when and where?[J]. Clin Kidney J, 2020, 14(1): 49-58. DOI: 10.1093/ckj/sfaa188.

39.Obermeyer Z, Emanuel EJ. Predicting the future-big data, machine learning, and clinical medicine[J]. N Engl J Med, 2016, 375(13): 1216-1219. DOI: 10.1056/NEJMp1606181.

40.李家宁, 熊睿彬, 兰艳艳, 等. 因果机器学习的前沿进展综述[J]. 计算机研究与发展, 2023, 60(1): 59-84. [Li JN, Xiong RB, Lan YY, et al. Overview of the frontier progress of causal machine learning[J]. J Comput Res Dev, 2023, 60(1): 59-84.] DOI: 10.7544/issn1000-1239.202110780.