Ability of Thromboelastography to Detect Hypercoagulability: A Systematic Review and Meta-Analysis.

Abstract

To review the current literature on the use of viscoelastic hemolytic assays, such as thromboelastography (TEG) and rotational thromboelastometry (ROTEM), during the perioperative period of patients and determine the ability of TEG and ROTEM to detect hypercoagulability and identify increased risk of the development of venous thromboembolism (VTE). PubMed, EMBASE, and Cochrane online databases were queried through February 11, 2018, by pairing the terms "thromboelastography," "viscoelastic hemostatic assays," and "rotational thromboelastometry" with "venous thromboembolism," "deep vein thrombosis," "pulmonary embolism," and "hypercoagulability." Inclusion and exclusion criteria were established to determine relevance and quality of data, of which 2.54% of initially identified studies met. Articles and citations were reviewed for relevance by 2 independent individuals following PRISMA guidelines as well as a quality assessment of data as established by Zaza et al. In studies that separated patients postoperatively by VTE development or no VTE development, data were pooled utilizing a modified DerSimmion and Laird random effects model. One thousand eight hundred ninety-three articles were assessed for eligibility, yielding 370 abstracts. Of the 370 abstracts, 35 studies were included, and of these, only 5 were included in the meta-analysis. Studies included postsurgical patients in a variety of surgical fields, encompassing a total of 8939 patients, with 717 thrombotic events reported. Elevated maximum amplitude (MA) was a statistically significant indicator of hypercoagulability across at least 1 perioperative time point in 17 (50%) of the articles reviewed, consisting of 6348 (72%) patients. The pooled mean MA value for defining hypercoagulability was greater than 66.70 mm. Using a prepublished value for hypercoagulability of 65 mm, the combined effect of MA on the development of VTE in postsurgical patients was determined to be 1.31 (95% confidence, 0.74-2.34, P = 0.175) and was 46% sensitive and 62% specific in predicting a postoperative VTE. Only 1 parameter, MA, was consistently used to both define hypercoagulability and be predictive of VTE after traumatic injury and surgical intervention; however, there remains a broad variability in the definition of hypercoagulability as determined by MA and thus limits its predictive ability. In addition, when hypercoagulability was measured throughout the perioperative period, TEG consistently demonstrated hypercoagulability starting on post-op day 1 (POD1). Diagnostic Level III. See Instructions for Authors for a complete description of levels of evidence.