Highly sensitive qualitative methods for serum choriogonadotropin (hCG): clinical specificity studies

We screened six highly sensitive kits, designed for serum pregnancy tests, for false-positive results. The two best were then evaluated more extensively. The "BETA-hCG MAIA-clone" (I) and the "TANDEM ICON" (II) kits gave only negative results for 100 sera from men at 5, 10, and 25 int. units/L (1st IRP). Of 100 serum specimens from post-menopausal women three and 10 were hCG positive by the II and the I reagents, respectively, but only at the 5 int. units (1st IRP) of hCG per liter level of sensitivity. At 10 and 25 int. units/L, all specimens were negative by both tests. The manufacturers of these kits recommend pregnancy testing only at the 25 int. units/L level of analytical sensitivity. By quantitative methods, hCG concentrations in the II positive samples ranged from 6 to 20 int. units/L (1st IRP) and lutropin concentrations were between 130 to greater than 150 int. units/L. The medical records of the corresponding patients did not support the presence of trophoblast or any other hCG-secreting tissues. During 15 months of routine use of the II reagents at an analytical sensitivity of 25 int. units/L (1st IRP) hCG for pregnancy testing (greater than 4000 serum specimens from pre-menopausal women), the staff has not reported to us any suspected false-positive findings.