Potential dosing errors using portable prothrombin time monitoring devices

Ronen Loebstein, Daniel Kurnik, Aharon Lubetsky, David Ezra, Hillel Halkin*

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review


Portable prothrombin time (PT) monitors facilitate the control of warfarin therapy. Few studies have compared the influence of using different monitors on dosage decisions. We determined the comparability of data generated by two portable PT monitors, Coaguchek S, (Roche Diagnostics Boehringer-Mannheim) and Hemochron Jr (International Technidyne Corporation Ltd.), with that of a reference laboratory. Simultaneous International Normalized Ratio (INR) measurements (portable monitor and laboratory) were performed in 193 consecutive patients receiving warfarin for at least 3 months. Agreement of measurements was assessed by both regression analysis and influence on dosage decisions in accordance with pre-defined criteria. The Coaguchek S versus laboratory INR regression line (n = 111; r2 = 0.88; P < 0.001) was close to the line of identity, while that of the Hemochron Jr (n = 82; r2 = 0.61; P < 0.001) was not. The overall proportion of dual INR measurements that fulfilled the clinical criteria of agreement was 90% for the Coaguchek S compared with 62% for the Hemochron Jr (P < 0.0001). For laboratory INRs 2.0-2.5, 2.6-4.0 and > 4.0, the proportions of portable measurements that satisfied the clinical criteria for the Coaguchek S versus the Hemochron Jr were 96 versus 63% (P < 0.001), 81 versus 45% (P < 0.04), and 67 versus 17% (P < 0.85), respectively. Warfarin dosing based solely on the portable devices would have resulted in unjustified dose increments in 22% of the patients with the Hemochron Jr device compared with 8% with the Coaguchek S monitor (χ2 = 4.43; P = 0.035). The Coaguchek S monitor provides measurements for INR values within the therapeutic range that agree well with the standard laboratory. The Hemochron Jr measurements result in different dosage adjustments even within the therapeutic range, but especially for INR values > 4.0. For both monitors, agreement of INR measurements with the standard decreases with increasing INR values.

Original languageEnglish
Pages (from-to)479-483
Number of pages5
JournalBlood Coagulation and Fibrinolysis
Issue number5
StatePublished - Jul 2003
Externally publishedYes


  • Accuracy
  • Clinical agreement
  • Portable prothrombin time monitors


Dive into the research topics of 'Potential dosing errors using portable prothrombin time monitoring devices'. Together they form a unique fingerprint.

Cite this