Summary
Four thromboplastin reagents were tested by 18 laboratories in Europe, North-America,
and Australasia, according to a detailed protocol. One thromboplastin was the International
Reference Preparation for ox brain thromboplastin combined with adsorbed bovine plasma
(coded OBT/79), and the second was a certified reference material for rabbit brain
thromboplastin, plain (coded CRM 149R). The other two thromboplastin reagents were
another rabbit plain brain thromboplastin (RP) with a lower ISI than CRM 149R and
a rabbit brain thromboplastin combined with adsorbed bovine plasma (RC). Calibration
of the latter two reagents was performed according to methods recommended by the World
Health Organization (W. H. O.).
The purpose of this study was to answer the following questions: 1) Is the calibration
of the RC reagent more precise against the bovine/combined (OBT/79) than against the
rabbit/plain reagent (CRM 149R)? 2) Is the precision of calibration influenced by
the magnitude of the International Sensitivity Index (ISI)?
The lowest inter-laboratory variation of ISI was observed in the calibration of the
rabbit/plain reagent (RP) against the other rabbit/plain reagent (CRM 149R) (CV 1.6%).
The highest interlaboratory variation was obtained in the calibration of rabbit/plain
(RP) against bovine/combined (OBT/79) (CV 5.1%). In the calibration of the rabbit/combined
(RC) reagent, there was no difference in precision between OBT/79 (CV 4.3%) and CRM
149R (CV 4.2%). Furthermore, there was no significant difference in the precision
of the ISI of RC obtained with CRM 149R (ISI = 1.343) and the rabbit/plain (RP) reagent
with ISI = 1.14. In conclusion, the calibration of RC could be performed with similar
precision with either OBT/79 or CRM 149R, or RP.
The mean ISI values calculated with OBT/79 and CRM 149R were practically identical,
indicating that there is no bias in the ISI of these reference preparations and that
these reference preparations have been stable since their original calibration studies
in 1979 and 1987, respectively.
International Normalized Ratio (INR) equivalents were calculated for a lyophilized
control plasma derived from patients treated with oral anticoagulants. There were
small but significant differences in the mean INR equivalents between the bovine and
rabbit thromboplastins. There were no differences in the interlaboratory variation
of the INR equivalents, when the four thromboplastins were compared.
