A single serum test for measuring early pregnancy outcome with high predictive value

Objectives: Current testing to determine a failing pregnancy requires two separate clinic visits to measure the hCG doubling rate. Diagnosing a failing pregnancy is often done in emergency departments where simplified and accelerated testing methods are needed. Here, we investigated byperglycosylated hCG (hCG-H) for predicting pregnancy failure. Design and methods: We studied two independent sets of patient samples collected in the early weeks of gestation. One set was urine samples, and the other was serum samples. In all cases, hCG and hCG-H were measured using automated chemiluminescence immunoassays. Concentrations of hCG and hCG-H were plotted on a scattergram, and levels in failing pregnancies were compared to those in continuing pregnancies. Results: Data indicated that a threshold level of hCG-H (13 mu g/L) in both serum and urine samples defined the concentration below where pregnancies were likely to fail. This cut-off corresponded to 73% detection of failures at a 2.9% false positive rate using serum and 75% detection at a 15% false positive rate using urine. Using an hCG cut-off that corresponded to the same false positive rates, hCG detected only 42% of failures using serum and 43% of failures using urine. Conclusions: Our data indicate that hCG-H provides a much more accurate single test than hCG for assessing pregnancy outcome. Compatible with the use of serum or urine samples, a single hCG-H test might provide simpler, faster, and more accurate results for predicting the progress of a pregnancy than standard hCG testing. (c) 2006 The Canadian Society of Clinical Chemists. All rights reserved.