Measuring partial thromboplastin-time. An international collaborative study.

A series of collaborative exercises on the partial thromboplastin-time (P.T.T.) test, involving over three hundred hospital centres in Britain and overseas, were performed in 1975. Lyophilised test plasmas were issued from the World Health Organisation Collaborating Centre for Anticoagulant Control Reagents to participants, together with a standardised reference P.T.T. reagent and a standard technique. Hospitals were asked to test the plasma samples with the standardised reagent and technique in parallel with their customary local P.T.T. reagent and method. The overall success-rate in detecting the intrinsic clotting abnormality in the eight abnormal test samples was higher with the standardised reagent and technique than with all other reagents. Furthermore, fewer hospitals obtained false positive results when the normal plasma sample was tested with the standardised method rather than with their usual routine reagents. An index was used to measure the success-rate of the P.T.T. reagents in correctly identifying the test plasmas as normal or abnormal. The eight test plasmas showed a varying degree of abnormality. A system of "weighting" was therefore introduced as the failure of a P.T.T. method to detect more severe defects was regarded as more serious. Although hospitals were unfamiliar with the standardised method, the results established its superiority over all other P.T.T. reagents included in the trials in sufficient numbers for analysis. Failures with commercial reagents may have been caused by insensitivity of the cephalin extracts or the unreliability of the manufacturers recommended techniques. Since the same laboratories obtained good results with the standardised method technical failure can be excluded.