Analytic validity of genetic tests to identify factor V Leiden and prothrombin G20210A

The objective of this study is to systematically review methods for detecting Factor V Leiden or prothrombin G20210A. English-language literature from MEDLINE, EMBASE, The Cochrane Library, the Cumulative Index to Nursing and Allied Health Literature, PsycInfo(c), 2000-December 2008. Studies assessed methods for detection of these mutations in at least 10 human blood samples and reported concordance, discordance, or reproducibility. Two investigators abstracted data on the sample selection criteria, test operators, DNA extraction, experimental test, reference standard, commercial instruments, concordance rates, explanation of any discordance, and whether discordance resolved after repetition. We assessed strength of the evidence using the GRADE criteria. We reviewed 7,777 titles and included 66 articles. The majority of the reviewed studies used PCR-RFLP or AS-PCR as the reference standard. The studies demonstrated that commercially available and precommercial tests have high analytic validity with all having greater than 99% concordance with the reference standard. With a few exceptions, discordance resolved with repetition of the test, suggesting operator or administrative errors were responsible for the discordant results. In the quality assurance studies, greater than 98% of laboratories demonstrated high, even perfect, accuracy when asked to diagnose a sample with a known mutation. The majority of errors came from a limited number of laboratories. Although not all methods may be accurate, there is high-grade evidence that genetic tests for the detection of FVL and prothrombin G20210A have excellent analytic validity. There is high-grade evidence that most, but not all, clinical laboratories test for FVL and prothrombin G20210A accurately.