An audit of the diagnostic accuracy of rotational thromboelastometry for the identification of hypofibrinogenaemia and thrombocytopenia during cardiopulmonary bypass.

We audited the diagnostic accuracy of ROTEM® (TEM Innovations, GmbH, Munich, Germany) measurements of hypofibrinogenaemia (fibrinogen <1.5 g/l) and thrombocytopenia (platelet count <100 x 10/l) in 200 adult non-transplant patients during cardiopulmonary bypass (CPB). Blood samples were obtained for FIBTEM (assay for the fibrin part of the clot), PLTEM (calculated platelet-specific component), and laboratory measurements simultaneously. Our thresholds for FIBTEM and PLTEM were A10 (clot firmness 10 minutes after clotting time) ≤8 mm, and <35 mm respectively. We also calculated the accuracy of smaller thresholds and earlier indices. We found that FIBTEM A10 ≤8 mm had low sensitivity (0.62) for hypofibrinogenaemia. With the 13% hypofibrinogenaemia prevalence in our sample, the positive predictive value (PPV) was 0.47. In contrast, FIBTEM A10 ≤8 mm had higher specificity (0.90) (negative predictive value [NPV] 0.94). Of the other FIBTEM values analysed, only A5 ≤6 mm had similar or superior accuracy. The PLTEM results were less encouraging (sensitivity 0.81, specificity 0.62). With our prevalence of thrombocytopenia (also 13%), the PPV was only 0.24. However, the NPV was high (0.96). Of the other PLTEM values analysed, only A5 <25 mm had similar or superior accuracy. These findings indicate that during CPB FIBTEM A10 ≤8 mm and PLTEM A10 <35 mm have greater accuracy in identifying the absence of hypofibrinogenaemia and thrombocytopenia respectively than their presence. On the basis of these results we would be reassured by FIBTEM A10 values >8 mm and PLTEM A10 values ≥35 mm, but would continue to use laboratory measurements for confirmation. We would not use FIBTEM A10 ≤8 mm or PLTEM A10 <35 mm values alone to guide replacement therapy unless clinical conditions warranted an immediate decision before laboratory measurements were available.