Systematic review of bleeding risk prediction models for patients with venous thromboembolism
10.3760/cma.j.cn115682-20211122-05263
- VernacularTitle:静脉血栓栓塞症患者出血风险预测模型的系统评价
- Author:
Qiao HE
1
,
2
;
Xiaofan DOU
;
Xiaomin CHEN
;
Xueliang SONG
;
Hongying YU
Author Information
1. 浙江省人民医院&
2. 杭州医学院附属人民医院骨科，杭州 310014
- Keywords:
Venous thromboembolism;
Bleeding;
Risk prediction model;
Systematic review
- From:
Chinese Journal of Modern Nursing
2022;28(18):2443-2448
- CountryChina
- Language:Chinese
-
Abstract:
Objective:To systematically analyze and evaluate the bleeding risk prediction model of patients with venous thromboembolism (VTE) , so as to provide a reference for nursing practice.Methods:PubMed, Web of Science, Embase, Cochrane Library, Wiley Online Library, CINAHL, Scopus, ProQuest, Clinicalkey, Wanfang, China National Knowledge Infrastructure, VIP databases were searched for bleeding risk prediction model of VTE patients. The retrieval time was from the establishment of the databases to August 19, 2021. Two researchers independently screened the literature according to the inclusion and exclusion criteria, then extracted the data and evaluated the quality of the literature.Results:Ten relevant literatures were included in this study, including 4 retrospective studies and 6 prospective studies. The area under the receiver operating characteristic curve of one model was greater than 0.7, and the C-index of four models were greater than 0.7. The prediction model contained many relevant factors, including bleeding history, age, anemia, tumor, antiplatelet therapy, gender, renal insufficiency, etc. In the evaluation of risk of bias and applicability, 10 studies had good applicability, but there were bias mainly due to insufficient number of events in the dependent variable, conversion of continuous variables into dichotomous variables and missing data were not reported. Conclusions:The research on the prediction model of bleeding risk in VTE patients is still in the development stage. In the future, risk prediction models with excellent performance in all aspects and low risk of bias can be developed and verified externally.
