OC73 SYSTEMATIC REVIEW OF THE PREDICTIVE ACCURACY OF POINT OF CARE TESTS OF COAGULOPATHY IN CARDIAC SURGERY

Abstract

Abstract Background: Coagulopathic haemorrhage is a common and serious complication of cardiac surgery. Timely diagnosis and treatment of coagulopathy is considered an important determinant of outcome. The aims of the review were to quantify and compare the predictive accuracy of POC tests in clinical use. Methods: Observational studies and randomized controlled trials that report measures of predictive accuracy for POC tests that are, or have been used clinically for the diagnosis and treatment of coagulopathic haemorrhage. Results: 56 studies recruiting a total of 9123 participants, met the inclusion criteria for the qualitative analysis and 15 studies were included for quantitative analysis. We have shown that the existing evidence is of insufficient quality to determine the predictive accuracy of point of care diagnostic tests. The point estimates of the Summary AUCROC analyses and the analysis of sensitivity and specificity estimates from individual studies suggested that: Viscoelastic tests had low sensitivity and high specificity, Platelet Function Tests had low sensitivity and moderate specificity, and combined Viscoelastic and Platelet Function tests had high sensitivity and low specificity. However there was great uncertainty around all of these estimates and the 95% margins of uncertainty included values less 0.5 in every analysis. Calculated diagnostic odd ratios also indicated poor predictive accuracy for all of the groups. Conclusions: Coagulopathic haemorrhage remains a common and severe complication of cardiac surgery. A systematic review of available evidence did not demonstrate predictive accuracy for point-of-care tests in clinical use for the diagnosis of coagulopathy in cardiac surgery.