A 'blind' study has been made to try to find out if it is possible to diagnose carriers of haemophilia. A group of 34 obligatory carriers of haemophilia were compared with 34 normal women. Levels of factor VIII activity, factor VIII-related antigen, factor V and ratio of factor VIII activity to factor VIII-related antigen were measured. In the carrier group the mean level of factor VIII activity and the mean level of the ratio of activity to antigen were each approximately half of those found in the normal women. The mean level of factor V was the same in both groups of women. By setting the lower limit of normal at the lowest level of the different factors found in the normal women, 12 out of 34 (35%) carriers could be distinguished on the basis of their factor VIII level alone; 24 out of 34 (71%) could be detected on the basis of the ratio of factor VIII activity to factor VIII related antigen and 25 out of 34 (73%) could be detected if both factor VIII activity and the ratio were taken into account. It is concluded that consideration of both the level of factor VIII activity and the ratio of factor VIII activity to factor VIII-related antigen is of some value in detecting carriers of haemophilia. The number of carriers detected (73%) in the present study is not as high as that found by other workers.

Summary. A repeated finding of national and international collaborative studies of standard factor VIII preparations has been that systematic differences exist between laboratories in their measurement of the relative activities of the same pairs of factor VIII preparations. A workshop meeting was held at the Oxford Haemophilia Centre during 23–26 November 1976 to investigate which of the possible sources of variation between laboratories were responsible. Participants from 16 British laboratories (nine using one‐stage assays and seven using two‐stage assays) performed a total of 273 valid assays using three freeze‐dried preparations of differing purity (a plasma, an intermediate and a high purity concentrate). The results of assay with each participant using their normal system established that, if the participants were a representative cross‐section, approximately one‐third of one‐stage laboratories would show a systematic difference from the overall mean of at least 16% with a corresponding figure for the two‐stage laboratories of 9%. Various features of the assay systems were then modified in a controlled series of experiments. The results showed conclusively that (i) differences between reagents accounted for most of the variation between laboratories, and (ii) the two‐stage assays were, on average, detecting relatively more activity in the more purified preparations than the one‐stage assays. The results also suggested that the use of buffer or citrate‐saline as opposed to haemophilic plasma for the initial dilution of concentrates did not affect the assay results.

SummaryA collaborative trial has been carried out under the auspices of the International Committee on Thrombosis and Haemostasis to compare the Bethesda and New Oxford methods of antibody assay. It was found that errors between laboratories were much greater than those within laboratories and each laboratory had a bias whereby it always rated samples high or low with respect to the other laboratories. However there was excellent agreement in the order in which laboratories ranked antibody samples and if a standard antibody sample could be provided there would be a significant improvement in numerical agreement between laboratories. On average, for this exercise, a result for a given sample in Bethesda units was 1.21 times the result in New Oxford units although it must be stressed that this ratio could vary from sample to sample.

Plasma coagulation factors were measured in twelve male insulin-dependent diabetics with no retinopathy, ten with background and ten with proliferative retinopathy and ten non-diabetics. Factor VIII pro-coagulant activities (VIII:C), ristocetin cofactor activities and factor VIII-related antigen concentrations (VIIIR:ag) were significantly related to the severity of diabetic retinopathy (P less than 0.025, trend test). The mean ratio of VIII:C/VIIIR:ag was lower in the diabetics with proliferative retinopathy than in the other groups of diabetics (P less than 0.05) or the controls (P less than 0.02). Concentrations of alpha 2 macroglobulin and alpha 1 antitrypsin were highest in diabetics with proliferative retinopathy (0.1 greater than P greater than 0.05, trend test) but mean prothrombin and activated partial thromboplastin times and mean concentrations of alpha 2 antiplasmin, plasminogen activator and antithrombin III were similar in all groups. Concentrations of the platelet-specific protein beta thromboglobulin, though higher in diabetics than controls (P less than 0.005), were not related to retinopathy. The plasma concentrations of coagulation factors did not correlate with creatinine clearance and there were no significant differences between groups in concentrations C-reactive protein; this suggests that the raised concentrations of coagulation factors in diabetics with retinopathy were not a result of associated nephropathy or an 'acute phase protein' response to diabetic tissue damage. Increased coagulation activity in diabetics may contribute to the development of retinopathy.

A repeated finding of national and international collaborative studies of standard Factor VIII preparations has been that systematic differences exist between laboratories in their measurement of the relative activities of the same pairs of Factor VIII preparations.A workshop meeting was held at the Oxford Haemophilia Centre (England) during 23rd-26th November 1976 to investigate which of the possible sources of variation between laboratories were responsible. Participants from 16 British laboratories (9 one-stage, 7 two-stage) performed a total of 273 assays using three freeze-dried preparations of differing purity (a plasma, an intermediate and a high purity concentrate). The results of assays with each participant using their normal system established that, if the participants were a representative cross-section, approximately one-third of one-stage laboratories would show a systematic difference from the overall mean of at least 16%, with a similar figure for the two-stage laboratories of 9%. Various features of the assay systems were then modified in a controlled series of experiments. The results showed conclusively that i) differences between reagents accounted for most of the variation between laboratories and, ii) the two-stage assays were, on average, detecting relatively more activity in the more purified preparations than the one-stage assays. The results also suggested that the use of buffer as opposed to haemophilic plasma for the initial dilution of concentrates did not affect the assay results.

In assays of factor VIII antibody the patient’s antibody - containing plasma is allowed to react with factor VIII for a given length of time, the residual factor VIII assayed and the level of antibody calculated in arbitrarily defined units. The several methods of assay used throughout the word vary with regard to the source of factor VIII, the length of time the antibody is allowed to react with the factor VIII, the method of assay of residual factor VIII and the definition of the unit of antibody. A comparison has been made of 3 methods of antibody assay, namely, the method of Biggs & Bidwell, the method of Rizza & Biggs and the “Bethesda method”, the aim being to try to find out to what extent the results on any given sample varied with the method and to try to identify the main causes of variation. The investigation shows that the 3 methods record markedly different levels of antibody in any one sample. This is partly due to the different criteria for defining the antibody unit but in addition assays in which whole normal plasma is used as a source of factor VIII substrate show significantly higher levels of antibody compared with assays in which a concentrate of factor VIII is used as substrate. These results and their implications will be discussed.