Using weighted harmonic mean for prediction of APTT in the mixing test

Background

The mixing test of activated partial thromboplastin time (APTT) is used for differentiating factor deficiency (FD), lupus anticoagulant (LAC), and acquired hemophilia A (AHA). However, the interpretation of the mixing test is not fully standardized.

Objectives

The aim of this study was to determine whether the weighted harmonic mean predicts the APTT in mixture of a mixing test samples and is useful for the differentiation of FD, LAC, and AHA.

Patients/methods

We examined 27 FD, 26 LAC, and 18 AHA samples. Harmonic means of APTT were calculated from the clotting times with and without 2 h incubation. We defined the index of harmonic mean (IHM) as the ratio of the actual APTT to the predicted APTT calculated by the harmonic mean. We defined IHM of the measured immediate after mixing samples and of delayed (after 2 h of incubation) measured samples as IHMi and IHMd respectively.

Results

Actual APTT and predicted APTT were correlated in the FD group. Both IHMi and IHMd in the FD group were equal or lower than 1.02, whereas those in the LAC group were higher than 1.02. In the AHA group, the IHMd was higher than 1.02, whereas half of the IHMi were equal or lower than 1.02. Time dependent inhibition evaluated by IHMd/IHMi was not observed in the LAC group, whereas it was observed in 77% of participants in the AHA group.

Conclusions

The harmonic mean was potentially useful in predicting APTT in the mixing test, and the IHM calculated from the predicted APTT had differentiation potency for FD and LAC, and for FD and AHA. IHM was also available for partial differentiation of LAC to AHA.