Precision of progesterone measurements with the use of automated immunoassay analyzers and the impact on clinical decisions for in vitro fertilization

Abstract

To compare the precision of progesterone measurements obtained with the use of immunoassays and of liquid chromatography-tandem mass spectrometry (LC-MS/MS). Comparative study. Academic, private practice, and in vitro fertilization (IVF) research centers. A total of 189 human serum samples were collected during controlled ovarian hyperstimulation and early pregnancy in women undergoing IVF. Serum progesterone pools (n = 10; 0.2-4 ng/mL) were sent to four laboratory centers that used four different automated immunoassay analyzers. Progesterone was measured by immunoassay in triplicate at three separate time points (n = 9 per pool) and by LC-MS/MS in triplicate once (n = 3 per pool). Inter- and intraassay coefficients of variation (CVs) of progesterone measurements were compared for each analyzer and LC-MS/MS. Progesterone measurements by immunoassay were highly correlated with those by LC-MS/MS. Only two analyzers had intraassay CVs <10% at all three experimental time points, and only two analyzers had an interassay CV <10%. Mean progesterone levels by the analyzers were different across multiple progesterone pools. Our results indicate that progesterone threshold measurements used for IVF clinical decisions should be interpreted cautiously and based on laboratory- and method-specific data. A validated progesterone standard incorporated into daily immunoassays could improve medical decision accuracy.