Accuracy of Point of Care Testing in Patients on Dual Anticoagulation During Initiation of Anticoagulation or Bridging,

Point-of-care testing (POCT) is widely used for monitoring of the international normalized ratio (INR) in patients on oral anticoagulation with a vitamin-K antagonist (VKA) and numerous clinical studies have assessed the accuracy of this method in comparison with INR results from venous blood samples analysed in the laboratory. There is however a paucity of clinical data to support the use of POCT in patients on dual anticoagulation with low molecular weight heparin (LMWH) and a VKA during initiation of anticoagulation or bridging after a surgical procedure.

To test the hypothesis whether therapeutic doses of LMWH interfere with INR measurements when using a POCT system during times of dual anticoagulation with LMWH and a VKA. To further investigate whether the effect is most pronounced once LMWH has reached peak plasma levels and less evident 10 hours and more after administration of LMWH.

We prospectively collected 160 consecutive venous blood samples from patients on therapeutic doses of LMWH - Tinzaparin (175 IU/kg once daily) and a VKA commonly warfarin for INR testing in our laboratory. At the same time all patients had their INR determined on capillary blood collected by finger prick using a CoaguChek XS Pro and INR test strips with the same lot number (Roche Diagnostics Ltd, UK).

60 blood samples were collected within 3–6 hours after administration of LMWH (group 1) and 100 samples were collected 10 hours or more after the last injection of LMWH (group 2). For each sample the dose and time of the last injection of LMWH was recorded along with the time of the venepuncture and the result of the capillary INR. To ensure that we had a wide variation in the plasma concentrations of LMWH we carried out anti-Xa testing on a cross-section of venous samples

The dosing advice for Warfarin was based on the INR result of the venous blood sample processed in the laboratory.

The correlation coefficient between the POCT INR and the laboratory INR was 0.98 in group 1 and 0.97 in group 2.

In the Bland Altman analysis for group 1 the mean 95% confidence interval (CI) was 0.03 (range+/− 1.96 SD: −0.26 to +0.32) and for group 2 the mean 95% CI was 0.00 (range −0.28 to +0.29). These results are comparable to results of our internal quality control between POCT INR and laboratory INR in patients on VKA alone with a mean 95% CI of −0.02 (range −0.26 to +0.29).

The mean INR was 1.8 by both methods in group 1 and 1.7 by both methods in group 2 and anti-Xa levels ranged from 0 to1.19 U/mL.

A variation in the result of the POCT INR and laboratory INR of 0.5 or greater is thought to affect dosing decisions for Warfarin. Such a variation was observed in 3% (2/60) in group 1 and 2% (2/100) in group 2.

There was good accuracy of the INR obtained with the POCT system used and this was not affected by the timing of the administration of LMWH in relation to testing.

No relevant conflicts of interest to declare.