A systematic review and meta-analysis of risk prediction models for post-thrombotic syndrome in patients with deep vein thrombosis

• Published in: Heliyon Vol. 9; no. 11; p. e22226
• Main Authors: Guo, Xiaorong, Xu, Huimin, Zhang, Jiantao, Hao, Bin, Yang, Tao
• Format: Journal Article
• Language: English
• Published: Elsevier Ltd 01.11.2023; Elsevier
• Subjects: Clinical prediction model; Deep vein thrombosis; Post-thrombotic syndrome; Deep vein thrombosis; Post-thrombotic syndrome; Clinical prediction model
• ISSN: 2405-8440; 2405-8440
• DOI: 10.1016/j.heliyon.2023.e22226

This systematic review and meta-analysis aimed to systematically evaluate the prediction models for the risk of post-thrombotic syndrome (PTS) in deep vein thrombosis (DVT) patients. This systematic review and meta-analysis was guided by the Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA). A systematic search on the following electronic database: PubMed/MEDLINE, EMBASE, and Cochrane Library, and Chinese databases such as WANFANG and CNKI was conducted to look for relevant articles based on the research question. The risk of bias for each studies included was carried out based on Prediction Model Risk of Bias Assessment Tool (PROBAST). We identified 10 studies that developed a total of 13 clinical prediction models for PTS risk in DVT patients, 3 models were externally validated, 2 models were temporally validated. The top 5 predictors were: BMI (N = 9), Varicose vein (N = 6), Baseline Villalta Score (N = 6), Iliofemoral thrombosis (N = 5), and Age (N = 4). The high risk of bias was from the analysis domain, which the number of participants and selection of predictors often did not meet the requirements of PROBAST. A random-effects meta-analysis of C-statistics was conducted, the pooled discrimination was C-statistic 0.75, 95%CI (0.69, 0.81). Among the 13 PTS risk prediction models reported in this study, no prediction model has been applied to clinical practice due to the lack of external validation. In the development of prediction models, most models were not standardized in data analysis. It is recommended that future studies on the design and implementation of prediction models refer to Transparent reporting of a multivariable prediction model for individual prognosis or diagnosis (TRIPOD) and PROBAST. [Display omitted]