Does thromboelastography predict postoperative thromboembolic events? A systematic review of the literature.

BACKGROUND Since thromboelastography (TEG) can detect hypercoagulable states, it is a potentially useful test for predicting postoperative thromboembolic complications. Therefore, we performed a systematic review of the literature to evaluate the accuracy of TEG in predicting postoperative thromboembolic events. METHODS PUBMED and EMBASE electronic databases were searched by two independent investigators to identify prospective studies involving adult patients undergoing operative procedures in which a TEG test was performed perioperatively and outcomes were measured by reference standards. The quality of included studies was assessed and measures of diagnostic test accuracy were estimated for each included study. RESULTS Ten studies (with a total of 1056 patients) were included in this analysis; however, only five reported measures of TEG test accuracy. The overall quality of the studies and level of diagnostic evaluation of the studies were highly variable, from poor to good. As there were variations in the definition of hypercoagulability, TEG methodology and patient characteristics, reference standards used and outcomes measured, a meta-analysis was not undertaken. The sensitivity and specificity ranged from 0% to 100% and 62% to 92%, respectively. The diagnostic odds ratio ranged from 1.5 to 27.7; area under the curve ranged from 0.57 to 0.91. Of the TEG variables, maximum amplitude seems to be the best parameter to identify hypercoagulable states and to predict thromboembolic events. CONCLUSIONS The predictive accuracy of TEG for postoperative thromboembolic events is highly variable. To determine if the TEG is a clinically useful screening test in high-risk surgical populations, more prospective studies are needed.