International Council for Standardization in Haematology field study evaluating optimal interpretation methods for activated partial thromboplastin time and prothrombin time mixing studies

Context.—
The prothrombin time (PT) and activated partial thromboplastin time (APTT) are screening tests used to detect congenital or acquired bleeding disorders. An unexpected PT and/or APTT prolongation is often evaluated using a mixing test with normal plasma. Failure to correct (“noncorrection”) prolongation upon mixing is attributed to an inhibitor, whereas “correction” points to factor deficiency(ies).

Objective.—
To define an optimal method for determining correction or noncorrection of plasma mixing tests through an international, multisite study that used multiple PT and APTT reagents and well-characterized plasma samples.

Design.—
Each testing site was provided 22 abnormal and 25 normal donor plasma samples, and mixing studies were performed using local PT and APTT reagents. Mixing study results were evaluated using 11 different calculation methods to assess the optimal method based on the expected interpretation for factor deficiencies (correction) and noncorrection (inhibitor effect). Misprediction, which represents the failure of a mixing study interpretation method, was assessed.

Results.—
Percentage correction was the most suitable calculation method for interpreting PT mixing test results for nearly all reagents evaluated. Incubated PT mixing tests should not be performed. For APTT mixing tests, percentage correction should be performed, and if the result indicates a factor deficiency, this should be confirmed with the subtraction III calculation where the normal pooled plasma result (run concurrently) is subtracted from the mixing test result with correction indicated by a result of 0 or less. In general, other calculation methods evaluated that performed well in the identification of factor deficiency tended to have high misprediction rates for inhibitors and vice versa.

Conclusions.—
No single method of mixing test result calculation was consistently successful in accurately distinguishing factor deficiencies from inhibitors, with between-reagent and between-site variability also identified.