Summary
In collaborative experiments in 199 laboratories, nine commercial thromboplastins,
four thromboplastins held by the National Institute for Biological Standards and Control
(NIBS & C), London and the British Comparative Thromboplastin were tested on fresh
normal and coumarin plasmas, and on three series of freeze-dried plasmas. One of these
was made from coumarin plasmas and the other two were prepared from normal plasmas;
in each series, one plasma was normal and the other two represented different degrees
of coumarin defect.
Each thromboplastin was calibrated against NIBS&C rabbit brain 70/178, from the slope
of the line joining the origin to the point of intersection of the mean ratios of
coumarin/normal prothrombin times when the ratios obtained with the two thromboplastins
on the same fresh plasmas were plotted against each other. From previous evidence,
the slopes were calculated which would have been obtained against the NIBS&C “research
standard” thromboplastin 67/40, and termed the “calibration constant” of each thromboplastin.
Values obtained from the freeze-dried coumarin plasmas gave generally similar results
to those from fresh plasmas for all thromboplastins, whereas values from the artificial
plasmas agreed with those from fresh plasmas only when similar thromboplastins were
being compared.
Taking into account the slopes of the calibration lines and the variation between
laboratories, precision in obtaining a patient’s prothrombin time was similar for
all thromboplastins.
