Accuracy of aPTT monitoring in critically ill patients treated with unfractionated heparin.

INTRODUCTION The anticoagulant effect of unfractionated heparin (UFH) is usually monitored by means of the activated partial thromboplastin time (aPTT). In critically ill patients, however, increased levels of acute phase proteins may decrease the accuracy of the aPTT, leading to inadequate UFH dosing. In these circumstances, the anti-Xa assay is recommended for monitoring. OBJECTIVE We aimed to analyse the accuracy of the aPTT for the monitoring of UFH dosing in critically ill patients. METHODS In critically ill patients treated with therapeutic doses of UFH, we compared aPTT levels with simultaneously measured anti-Xa levels as the gold standard. Sensitivity and specificity of the aPTT were determined for different cut-off points, receiver operating characteristic (ROC) curves were constructed and their areas under the curve (AUCs) were calculated. RESULTS A total of 171 paired blood samples from 58 patients were analysed. Concordant aPTT and anti-Xa values were observed in 108 (63.2%) data pairs. In 33 data pairs (19.3%) the aPTT was discordantly high and in 30 data pairs (17.5%) discordantly low. The sensitivity of the aPTT in detecting UFH underdosing and overdosing was 0.63 and 0.37, respectively. When considering alternative thresholds, ROC curves for underdosing and overdosing had AUCs of 0.71 and 0.81, respectively. CONCLUSION In this small cohort of critically ill patients, the aPTT was accurate in 63.2% of the blood samples. Its sensitivity to detect UFH underdosing and overdosing was low (0.63 and 0.37, respectively). We conclude that in critically ill patients, the aPTT is not accurate enough to detect UFH underdosing and overdosing.