Use of viscoelastic tests to predict clinical thromboembolic events: A systematic review and meta-analysis

Abstract

We aimed to assess whether whole-blood viscoelastic tests are useful to identify patients who are hypercoagulable and at increased risk of thromboembolism. Two investigators independently analyzed studies in the MEDLINE, EMBASE, and Cochrane controlled trial register databases to determine the ability of viscoelastic tests to identify a hypercoagulable state that is predictive of objectively proven thromboembolic events. Forty-one eligible studies, including 10,818 patients, were identified and subject to meta-analysis. The majority of the studies (n = 36, 88%) used the maximum clot strength to identify a hypercoagulable state which had a moderate ability to differentiate between patients who developed thromboembolic events and those who did not (area under the summary receiver operating characteristic [sROC] curve = 0.70, 95% confidence interval [CI]: 0.65-0.75). The pooled sensitivity, specificity, and diagnostic odds ratio to predict thromboembolism were 56% (95%CI: 44-67), 76% (95%CI: 67-83), and 3.6 (95%CI: 2.6-4.9), respectively. The predictive performance did not vary substantially between patient populations, and publication bias was not observed. Current evidence suggests that whole-blood viscoelastic tests have a moderate ability to identify a variety of patient populations with an increased risk of thromboembolic events and can be considered as a useful adjunct to clinical judgment to stratify a patient's risk of developing thromboembolism.