A Review of beta2 -glycoprotein-l antibody testing results from a peer-driven multilaboratory quality assurance program.

We evaluated the results of lgG beta2-glycoprotein-I (B2GPI) antibody assays in a multilaboratory setting by analyzing data from an external quality assurance program for the 2003 through 2005 cycles for 27 serum samples, including quantitative IgG-B2GPI values and qualitative interpretation and grading (ie, negative or positive; grade of positivity), according to method type and in conjunction with clinical data. We report high interlaboratory variation in numeric IgG-B2GPI results, comparable to that reported for IgG anticardiolipin antibody (aCL) testing, and some method-based variation. For example, interlaboratory coefficients of variation for IgG-B2GPI were more than 50% in 19 samples (70%). For qualitative reporting, there was generally better consensus than previously reported for semiquantitative IgG-aCL testing; although 100% consensus occurred for only 11 samples (41%), more than 90% of laboratories agreed for 19 samples (70%). In some cases, laboratory findings (negative or positive IgG-B2GPI) did not agree with clinical information. Despite the lack of formal standardization for IgG-B2GPI testing compared with IgG-aCL, there seems to be better cross-laboratory consensus. Improvement in standardization of these assays is still required to improve interlaboratory and intermethod concordance of results and interpretation between laboratories and the clinical usefulness of IgG-B2GPI testing.

A multilaboratory peer assessment quality assurance program-based evaluation of anticardiolipin antibody, and beta2-glycoprotein I antibody testing.

We evaluated the performance of anticardiolipin (aCL) and beta2-glycoprotein I (beta2-GPI) antibody assays through a large external quality assurance program. Data from the 2002 cycle of the Royal College of Pathologists of Australasia Quality Assurance Program (RCPA QAP) were analyzed for variation in reported numerical values and semiquantitative results or interpretations according to method type or group and in conjunction with available clinical data. High interlaboratory variation in numerical results and notable method-based variation, combined with a general lack of consensus in semiquantitative reporting, continues to be observed. Numerical results from cross-laboratory testing of 12 serum samples (for immunoglobulin G [IgG]-aCL, IgM-aCL, and IgG-beta2-GPI) yielded interlaboratory coefficients of variation (CVs) that were higher than 50% in six of 12 (50%) specimens for IgG-aCL, and 12 of 12 (100%) specimens for IgM-aCL and IgG-beta2-GPI. Semiquantitative reporting also varied considerably, with total (100%) consensus occurring in only four of 36 (11%) occasions. General consensus (where > 90% of participating laboratories agreed that a given serum sample gave a result of either negative or positive) was only obtained on 13 of 36 (36%) occasions. Variation in results between different method types or groups were also present, resulting in potential biasing of the RCPA QAP-defined target results by the large number of laboratories using the dominant aCL assays. Finally, laboratory findings frequently did not agree with the available clinical information. In conclusion, in a large proportion of specimens from the 2002 RCPA QAP cycle, laboratories could not agree on whether a serum sample tested was aCL-positive or aCL-negative, or beta2-GPI-positive or beta2-GPI-negative. Despite prior attempts to improve the standardization of testing and reporting practices, laboratory testing for aCL and anti-beta2-GPI still demonstrates significant interlaboratory and intermethod variation, which needs to be taken into account for the clinical interpretation of test results, especially those from different laboratories.