Identifying carriers of mild haemophilia.

The problems of carrier identification in mild haemophilia were examined by studying a large kindred transmitting VIII:C levels which average 17 u/dl in affected males. Fifteen obligatory carriers and 13 normal women from this kindred were used as reference groups to produce a set of linear discriminants. The utility of these discriminants in identifying carriers of mild haemophilia was compared with that of a similar set of discriminants which had been prepared for use with carriers of severe haemophilia, each set of discriminants being tested against both data bases. It was found that both sets of discriminants had large error rates when applied to the mild data base, and that both had much smaller error rates when applied to the severe data base. This outcome resulted from the greater overlap between the factor VII-related activities of mild carriers and normal women than the overlap of these activities between severe carriers and normal women. Our conclusions were that (i) correct classification of many but not all women from families of mild haemophilia would be difficult, (ii) discriminants prepared using obligatory carriers of severe haemophilia were not more inefficient in mild haemophilia than those prepared using obligatory carriers of mild haemophilia, (iii) a single set of discriminants prepared from reference groups of normal women and obligatory carriers of severe haemophilia may be used for identifying potential carriers of all grades of haemophilia, and (iv), statistically speaking, VIIIR:Ag assays provide a limited amount of information for this purpose.