Send to

Choose Destination
Proc Soc Exp Biol Med. 1987 Feb;184(2):201-5.

Verification of early pregnancy tests in a multicenter trial.


Tests for the diagnosis of early pregnancy have been available since 1974. However, no studies have systematically verified the accuracy of routine clinical laboratories in measuring human chorionic gonadotropin (hCG) prior to the time that pregnancy is clinically evident. We have conducted such a study in association with the NICHD-funded Diabetes in Early Pregnancy Study (DIEP). The purpose of this study was to elucidate the etiology of malformations in pregnancies complicated by diabetes mellitus, which probably occurs within the first few weeks of pregnancy, and therefore uniformity of pregnancy testing was necessary among the five centers to find an association of a teratogen at the time of organogenesis. We confirmed that routine clinical laboratories, in fact, could measure accurately hCG at the time of the missed menses; however, detection was not necessarily possible prior to that time. We conclude that in order to assure accurate diagnosis of early pregnancy, tests should ordinarily be delayed until time of the missed menses. When the test is used at this time, it is a reliable tool for early pregnancy testing and thus can be used to resolve questions relating to early pregnancy pathophysiology.


This paper describes the validation procedure of early pregnancy tests within 2 days of missed menses by 6 different university clinical laboratories assaying for human chorionic gonadotropin (hCG) by radioimmunoassay (5) or radioreceptor assay (1). Test materials were known amounts of hCG added to male serum or plasma, divided into 2 aliquots, and labeled "research" or with a fictitious patient name. Concentrations of hCG were 0, 150, 300, 600, 1200, 2400 and 4800 mIU/ml. In a 2nd test, concentrations ranged from 10 to 50 mIU/ml. In the 1st verification series, there was general agreement, with only 1 center finding a false positive pregnancy. In the 2nd series, all laboratories identified correctly the serum without hCG. However, the unknown with a quantity of hCG in the range of LH cross-reactivity was called positive by 2 labs. The other 4 called the result "equivocal." All 6 laboratories had adequate agreement between "research" and pseudo-patient samples. All laboratories could detect hCG at levels compatible with those observed at the time of a missed period, less than 200 mIU/ml. Even large inter-assay variation in 1 center failed to affect the ability of that laboratory to detect low levels of hCG. 2 laboratories did not measure hCG levels above 1200 mIU accurately, not a problem in diagnosing early pregnancy, but a possible limitation in following multiple gestation or hydatidiform mole. This study demonstrates the importance of verification procedures before clinical studies that need an accurate early pregnancy diagnosis.

[Indexed for MEDLINE]

Supplemental Content

Loading ...
Support Center