Abstract

AIMS: To compare the International Normalised Ratio (INR) obtained directly with the two types of WHO plain International Reference Preparation for thromboplastin in patients treated with coumarin. METHODS: Prothrombin times were performed in parallel at four centres using WHO human plain IRP (BCT/253) and rabbit plain IRP (RBT/79). Sixty patients and 20 normal controls were tested at each centre. Differences in INR among the centres were assessed by one factor, analysis of variance. The bias for each centre was assessed by the t test. RESULTS: At all four centres higher INRs were consistently found with the rabbit plain reagent. Two of the centres showed significantly greater bias. CONCLUSIONS: There was a small but significant difference in INR results obtained directly with these two reference reagents at all four centres (mean 7.35%). This in part may result from the different responsiveness of the two IRP to the coumarin defect or to imprecision of the original ISI calibrations of the two plain WHO IRP. The findings support the adoption of a single master IRP, in accord with WHO recommendations, which would resolve the present anomalous situation.