Survey of prothrombin time in National External Quality Assessment Scheme exercises (1980-87).

National External Quality Assessment Scheme surveys on the prothrombin time test carried out in hospitals in the United Kingdom have been performed at regular intervals since 1972. Performance has been assessed by comparing observed variability between hospitals with that predicted by a statistical model. The model was based on results from 53 survey plasmas issued between 1980 and 1987. These showed a linear correlation between logarithms of mean and standard deviation of reported ratios. Precision improved until the human brain thromboplastin, Manchester Comparative Reagent, was withdrawn in January 1986. There then followed a pronounced overall deterioration which, by October 1987, had not corrected to the levels achieved by 1985. When the recent results from 1986-87 were analysed according to Quick test reagent only one reagent (ISI 1.1) showed an improvement in precision. Performance of the other Quick test reagents, all with higher ISI values, had not regained the standards of precision previously achieved by the human brain reagent.