Background: Unfractionated heparin (UFH) has been in clinical use for more than half a century. Monitoring UFH treatments is mandatory due to large inter- and intra-individual variations in its anticoagulant activity, with a risk of thrombosis in case of under-dosing and a risk of hemorrhage in case of over-dosing. Laboratory monitoring of UFH treatments is usually based on the prolongation of activated partial thromboplastin time (aPTT), or on the measurement of plasma anti-Xa activity. As UFH present in the patient's blood sample could be neutralized by platelet factor 4 (PF4) released by in-vitro activated platelets, it is recommended to perform the tests aimed at evaluating its anticoagulant activity as soon as possible after blood collection. Actually, the current guidelines recommend a maximum delay of 2 hours between blood sampling and testing for anti-Xa activity or aPTT prescribed for monitoring treatments by UFH, when blood is collected into citrated tubes. As such a short delay could be an issue, particularly for multisite centres, we evaluated the potential impact of a longer delay on test results.
Design of the study: For that purpose, 2 citrated evacuated tubes containing 0.109 M tri-Na citrate (1 vol./9 vol.) were collected from patients on UFH: one was centrifuged and tested within 1h after blood collection (T1h) and one was stored for 4h at room temperature (+22°C) before being centrifuged and immediately analyzed (T4h).
Methods: Anti-Xa activity was evaluated using 2 reagents: Biophen Heparin LRT (Hyphen Biomed, Neuville-sur-Oise, France) and HemosIL Liquid Heparin (Instrumentation Laboratory, IL, Bedford, MA, USA). aPTT was evaluated using the HemosIL SynthASil reagent (IL). All assays were automated on an ACL TOP 700 CTS (IL). As the distributions of the data were not found to be normal, anti-Xa and aPTT results were expressed as the median values (with ranges), and comparison of test results obtained at T1h and T4h performed using the Wilcoxon test. Test results were also compared according to Bland-Altman.
Results: A total of 123 paired tubes were investigated. Analytical comparison of anti-Xa activity demonstrated a significant decrease (p<0.0001) after a 4 h-storage at room temperature (T4h) vs. a <1h-delay (T1h), for the two evaluated reagents (Table). The mean bias between test results obtained at T4h and T1h, evaluated according to Bland and Altman, was <0.05 IU/mL (in absolute value) for the two reagents, and identical for anti-Xa activities below or above 0.50 IU/mL. Such a value was below the imprecision of the techniques. There were 12 cases of discrepancy as whether test results were within the therapeutic range (0.30 - 0.70 IU/mL) or not when evaluated at T4h vs. T1h using the Hyphen reagent (9.8%), and 12.9% with the IL reagent. In most cases, these discrepancies were found for anti-Xa activities close to the lower limit of the therapeutic range, and would not have induced any change in the management of anticoagulation in these patients. APTT was significantly shortened (p<0.0001) after a 4h- vs. a 1h-storage at room temperature, with a mean bias of -8.1 sec, corresponding to a -0.25 decrease in ratio [relative change of -12.1% (95%CI= -34.8 +10.6)]. If considering the aPTT therapeutic range corresponding to anti-Xa activity between 0.30 and 0.70 IU/ml, 29 cases of discordant test results (23.6%) were observed that could have induced changes in the management of anticoagulation in 10% of the patients. Finally, the concordance whether test results measured at T4h and T0 were within or outside the therapeutic range was excellent for anti-Xa activity (kappa=0.813) and good for aPTT (kappa=0.661).
Conclusions: These results suggest that extending to 4h the delay between blood sampling and measurement of anti-Xa activity prescribed for monitoring UFH treatments when blood is collected into citrated tubes is acceptable and safe, at least when evaluated with the two tested reagents. As change in aPTT results was found to be of a greater order of magnitude and more relevant, it could be justified recommending a shorter delay.
No relevant conflicts of interest to declare.
