INR calibration of Owren-type prothrombin time based on the relationship between PT% and INR utilizing normal plasma samples

Abstract

SummaryProthrombin time (PT) is clinically important and is used to monitor oral anticoagulant therapy. To obtain PT results in international normalized ratio (INR), the current standardization procedure is complex and involves reference reagents.The PT of diluted plasma samples can be determined with a combined thromboplastin (the Owren-type procedure), but not necessarily with a plain thromboplastin (the Quick-type procedure). Owren-type PT procedures can therefore, as an alternative to the INR calibration, be calibrated with diluted normal plasma to give PT results in percent of normal PT activity (PT%).The present study explored if a plasma-based calibration of an Owren-type PT procedure can be used to obtain results in INR. The approach was to establish a relationship between PT% and INR by multi-center analysis of 365 samples from healthy individuals and patients on warfarin treatment. INR values were obtained by manual Quick-type reference procedure and PT% values by various automated Owren-type procedures. A relationship INR = (1/PT% + 0.018)/0.028 was found. A calibration procedure, based on the relationship, was investigated. Calibrators were the median PT of 21 normal plasma at dilutions representing 100%, 50%, 25%, 12.5% and 6.25% of normal PT activity. These were assigned INR values of 1.00, 1.36, 2.07, 3.05 and 6.36. Calibration of various Owren-type assays was repeatedly performed by 5 expert laboratories during 3 consecutive years. The INR values of certain lyophilised or frozen control plasmas were determined. The frozen control plasmas had externally assigned INR values according to WHO guidelines. Within the laboratory, CV was typically below 3%. No appreciable difference among the results of the different laboratories or the three assay occasions was found. Externally assigned and INR values were essentially identical to those found. These and other results indicated that the calibration procedure was reproducible, precise and accurate. Thus, an Owren-type PT assay can be calibrated with normal plasma samples to give results in INR and the investigated calibration procedure can be proposed for this purpose.