Therapeutic concordance of two portable monitors and two routine automatic oral anticoagulant monitoring systems using as reference the manual prothrombin time technique.

Two models of capillary blood prothrombin time (PT) monitoring systems were evaluated for analytical performance and then compared with two routine PT systems using the reference manual technique and a high-sensitivity thromboplastin. Two sets of 60 and 80 plasmas were analyzed from anticoagulated patients stabilized over 3 months in an INR range 2-3.5 for therapy. Capillary PT determination was performed in two portable monitors, CoaguChek S and CoaguChek PT (Roche Diagnostics), and plasma automatic methods were Neoplastine/STA (Diagnostics Stago) and PT-FibrinogenHsPlus/ACL7000 (Instrumental Laboratories). Thromboplastin Bilbao (TBi), an in-house high-sensitivity rabbit thromboplastin (ISI=1.08), recommended as the reference reagent by an External Spanish Oral Anticoagulant Quality Assessment, was used in the PT manual technique. The two monitors' coefficients of correlation with the reference system were 0.74 for CoaguChek S and 0.81 for CoaguChek PT. The automatic routine systems showed a correlation of 0.92 (Neoplastine/STA) and 0.91 (PT-FbHsPlus/ACL7000). Clinical agreement expressed as the percentage of simple correlation ranged between 75.0% (CoaguChek S) and 88.9% (Neoplastine/STA). The systems having the best kappa index with the manual technique were CoaguChek PT (71.9%) and the Neoplastine/STA system (73%). The routine PT management systems exhibited better correlation and percentage of concordance when using the TBi/manual technique than did the portable monitors, which moreover performed unequally in this regard.