Back

A comparison between routine plasma INRs and fingerstick whole blood INRs determinations, using three portable self-testing devices (Coagucheck®, Pro-Time®, Gem-PLC®) and a near-patient whole blood device (Thrombotrack®)

Abstract number: CD018

De Monte* P., Filippucci† E., Boschetti† E., Trottini† S., Cavalieri† G., Gresele† P.

†Italy *University of Perugia, Italy;

Background  

Oral anticoagulant therapy is conventionally monitored by the International Normalized Ratio (INR) on plasma samples. In alternative, capillary blood monitoring has been used for years by adopting the Thrombotest® reagent that, differently from the prothrombin time (PT), detects oral anticoagulant-induced deficiency of factor IX. More recently, several portable monitors for INR determination on capillary whole blood have been introduced for the self-testing of oral anticoagulation. No systematic simultaneous comparison of the Thrombotest® near-patient and of portable self-testing whole blood devices, has been performed so far, to the best of our knowledge.

Objective  

To compare the accuracy of four whole blood coagulation monitors (Thrombotrack®, Hyland Baxter Immuno; CoaguCheck S®, Roche; GEM PLC®, IL; Pro-time®, IL) in reference to standard PT (IL Recombiplastin) for INR determination on a wide range of anticoagulation levels.

Materials and methods  

Dual INR measurements were performed in patients on long-term oral anticoagulation, stabilized for at least 6 months both from finger capillary blood, by one of the four monitors, and from venous blood by the plasma PT. The accuracy of the whole blood coagulation monitors was assessed by evaluating, in comparison to laboratory PT: correlation and agreement; the absolute difference in INR; the proportion of whole blood monitor results within 0.5 INR units of laboratory PT.

Results  

c2 = 26.8, P < 0.001, #P < 0.001 (Table 1).

Discussion  

Whole blood monitors showed a good correlation with plasma PT. However, significant differences concerning agreement and the mean of differences as compared to plasma PT exist between the various whole blood monitors. In particular, some of the whole blood monitors showed a very low concordance and agreement with values in the high therapeutic range.

Table 1  

Parameter compared
to plasma INR
Thrombotest
(N = 230)
Coagucheck
(N = 167)
GEM PCL
(N = 99)
Pro-Time
(N = 97)
Correlation (r2) – all INRs0.800.770.610.68
Concordance limits (CL; 95% CI)0.63; +0.54-0.34; +1.22-1.7; +1.17-0.64; +0.54
Agreement (% of results within CL)83808172
Mean of differences (INR monitor – INR lab)-0.030.44#-0.55#-0.05
Percentage of results within 0.5 INR units
    All INRs*79655170
    INR <291909785
    INR 2–388655282
    INR >36243655
*c2 = 26.8, P < 0.001.
#P < 0.001.

To cite this abstract use the following format:

Journal of Thrombosis and Haemostasis 2003; 1 Supplement 1 July: abstract number

Session Details

Date: 14/07/2003
Time: 09:30-11:00
Session name: TTP/HUS
Subject: Clinical trials: heparins/LMWH
Location: Hall 3
Presentation type: Symposium
Back to top