The international normalized ratio (INR) for monitoring warfarin therapy: reliability and relation to other monitoring methods.

Objective: To enhance understanding of the reliability of the international normalized ratio (INR) for monitoring warfarin therapy and its relation to other monitoring techniques.

Design: Prospective cohort study.

Setting: A university hospital.

Patients: 79 patients attending an anticoagulation clinic.

Measurements: International normalized ratios obtained with a portable capillary monitor (Coumatrak) and the following from a simultaneous plasma sample: INRs from prothrombin times done with six thromboplastins, prothrombin-proconvertin (P&P) test activity, specific prothrombin activity, and native prothrombin antigen.

Results: Converting to INRs failed to standardize prothrombin time results obtained with high- and low-sensitivity thromboplastins. Coumatrak INRs correlated best with INRs obtained with high-sensitivity thromboplastins. The INR range of 2.0 to 3.0 corresponded to a P&P range of 30% to 13%, a native plasma prothrombin antigen range of 56 to 24 micrograms/mL, and a specific prothrombin activity range of 43% to 21%.

Conclusions: Low-sensitivity thromboplastins may give erroneously high INRs in the upper therapeutic range. Plasma prothrombin times should be done with a high-sensitivity thromboplastin, particularly in patients maintained at the upper limit of the therapeutic range. An INR so obtained correlated well with an INR obtained with a portable capillary blood monitor.