A cooperative study for the detection of the carrier state of classic hemophila

To determine the specificity and sensitivity of current technics for detecting carriers of classic hemophilia, three investigators simultaneously tested obligate carriers and noncarriers for immunologic and procoagulant factor VIII activity. Overall correct classification ranged from 72% (36 of 50) to 94% (47 of 50). The maximum accuracy obtained with the same linear-discriminant-function method on all data was 90% (26 of 29) in detecting carriers without misclassifying normal persons as carriers (none of 21). Lowest accuracy by the same technic was 66% (19 of 29) carrier detection while misclassifying 19% (4 of 21) normal persons. Precision of testing for both factor VIII activity and antigen was high (standard deviations from 0.004 to 0.026 on a log scale). Differences between participants seemed related to laboratory technics rather than to statistical methods. The factor VIII activity/antigen measurement is a valid technic for detecting the carrier state of hemophilia A.