Differences in Serum Human Chorionic Gonadotropin Rise in Early Pregnancy by Race and Value at Presentation

Abstract

Objective: To assess whether variation in serum human chorionic gonadotropin (hCG) measures, used to assess early gestation viability, are associated with differences in clinical presentation and patient factors.

Method: This retrospective cohort study included 285 women with first-trimester pain and bleeding and a pregnancy of unknown location for whom a normal intrauterine pregnancy was ultimately confirmed. Serial samples were collected at three U.S. sites and hCG changes were analyzed for differences by race, ethnicity, and clinical factors. A nonlinear, mixed-effects regression model was used assuming a random subject shift in the time axis.

Results: The hCG rise in symptomatic women with ongoing intrauterine pregnancy differs by patient factors and level at presentation. The 2-day minimum (first percentile) rise in hCG was faster when presenting hCG values were low and slower when presenting hCG value was high. African American women had a faster hCG rise (P<.001) compared with non-African American women. Variation in hCG curves was associated with prior miscarriage (P=.014), presentation of bleeding (P<.001), and pain (P=.002). For initial hCG values of less than 1,500, 1,500-3,000 and greater than 3,000 milli-international units/mL, the predicted 2-day minimal (first percentile) rise was 49%, 40%, and 33%, respectively.

Conclusion: The rise of hCG levels in women with viable intrauterine pregnancies and symptoms of potential pregnancy failure varies significantly by initial value. Changes in hCG rise related to race should not affect clinical care. To limit interruption of a potential desired intrauterine pregnancy, a more conservative "cutoff" (slower rise) is needed when hCG values are high.

Clinical trial registration: ClinicalTrials.gov, https://clinicaltrials.gov, NCT00194168.

Figure 1

Comparison of regression curves (assuming a quadratic model), subdivided by race, for estimating hCG rise in a viable intrauterine pregnancy. Plots show untransformed hCG values with logarithmic scaling. Triangles indicate the curve for African American women and circles indicate the curve for non-African American women. Dotted lines indicate the 95% confidence intervals.

Figure 2

Comparison of regression curves (assuming a quadratic model), subdivided by significant patient factors, for estimating hCG rise in a viable intrauterine pregnancy. Plots show untransformed hCG values with logarithmic scaling. Triangles indicate the curve for (A) prior miscarriage, (B) bleeding, (C) pain, and (D) age older than 34 years. Circles indicate the absence of the relevant condition. Dotted lines indicate the 95% confidence intervals.