Abstract

The management of post-operative bleeding in paediatric cardiac surgery is difficult because the underlying haemostatic changes have been poorly studied. The thromboelastogram (TEG) provides a real-time, functional measure of haemostasis, quantifying clot formation, strength and lysis. We investigated the use of serial intra-operative TEG to assess perioperative haemostatic changes and as an adjunct to demographic and laboratory variables for the prediction of bleeding following cardiopulmonary bypass (CPB) surgery. Ethical committee approval was obtained to study fifty-one children, median (interquartile) age 6.8 months (0.5 to 10.5) prospectively through CPB and for the first 24 hours after intensive care unit admission. Significant post-operative blood loss was defined as more than 10ml/kg in the first four post-operative hours. TEG readings and traditional coagulation parameters were measured throughout CPB. Forward stepwise logistic regression analysis was used to predict bleeding.

The incidence of bleeding was 37% (19/51), with a mean 4-hour blood loss of 24 ml/kg. Both groups showed abnormalities in all TEG parameters (apart from lysis) at the end of CPB, which were more marked among those who bled (all p<0.05).

Hyperfibrinolysis was not seen. Addition of heparinase to the TEG samples taken after protamine administration demonstrated a small but consistent heparin effect for all TEG parameters only among patients who bled, however this group exhibited comparable protamine:heparin ratios to those who did not (p = 0.22). Significant multivariate predictors of bleeding included fibrinogen level at induction of anaesthesia; occurrence of circulatory arrest; and two TEG parameters taken at the end of CPB: “k” (the time to maximal clot formation) and “angle” (the rate of clot formation). The final model predicted bleeding well, producing an area under the receiver operating characteristic curve of 0.855 (95% CI 0.750 to 0.960).

In summary, the aetiology of bleeding after paediatric cardiac surgery appears to be due to inadequate clot formation rather than fibrinolysis. TEG readings taken at the end of CPB, in conjunction with laboratory and demographic variables may provide a useful predictor of post-operative bleeding.

Author notes

Corresponding author