OBJECTIVE: To determine whether a combination of serum and urine biomarkers drawn from symptomatic pregnant women will help early differentiation of viable from nonviable pregnancies. METHODS: We conducted a prospective cohort study of 220 women who presented in the first trimester of pregnancy with complaints of pain, cramping, bleeding, or spotting. Serum samples for progesterone, inhibin A, and hCG, and urine beta-core hCG, were collected at presentation. To evaluate whether those biomarkers could predict viable and nonviable outcomes in pregnancy, we used likelihood ratios to compare operating characteristics of single and multiple biomarker strategies. RESULTS: Of 220 pregnancies studied, 98 were viable and 122 nonviable. Among single biomarkers, progesterone alone appears to have the greatest utility (area under the receiver operator characteristic curve = 0.923). Among dual-biomarker strategies, progesterone plus hCG and progesterone plus inhibin A improved specificity but not sensitivity. At 95% sensitivity, the combination of progesterone and hCG improved specificity from 0.29 to 0.66 (improvement = 0.37 [95% confidence interval 0.23, 0.52]). A triple-biomarker combination did not show substantial improvement over the dual-biomarker strategy. Also, combinations that used urine beta-core hCG did not improve diagnostic accuracy. CONCLUSION: Serum progesterone appeared to be the single most specific biomarker for distinguishing viable from nonviable pregnancies. When a dual-biomarker strategy was applied, combining serum progesterone with hCG, specificity improved significantly, which suggests that a multiple biomarker strategy might help distinguish viable from nonviable pregnancies in early gestation.
OBJECTIVE: To determine whether a combination of serum and urine biomarkers drawn from symptomatic pregnant women will help early differentiation of viable from nonviable pregnancies. METHODS: We conducted a prospective cohort study of 220 women who presented in the first trimester of pregnancy with complaints of pain, cramping, bleeding, or spotting. Serum samples for progesterone, inhibin A, and hCG, and urine beta-core hCG, were collected at presentation. To evaluate whether those biomarkers could predict viable and nonviable outcomes in pregnancy, we used likelihood ratios to compare operating characteristics of single and multiple biomarker strategies. RESULTS: Of 220 pregnancies studied, 98 were viable and 122 nonviable. Among single biomarkers, progesterone alone appears to have the greatest utility (area under the receiver operator characteristic curve = 0.923). Among dual-biomarker strategies, progesterone plus hCG and progesterone plus inhibin A improved specificity but not sensitivity. At 95% sensitivity, the combination of progesterone and hCG improved specificity from 0.29 to 0.66 (improvement = 0.37 [95% confidence interval 0.23, 0.52]). A triple-biomarker combination did not show substantial improvement over the dual-biomarker strategy. Also, combinations that used urine beta-core hCG did not improve diagnostic accuracy. CONCLUSION: Serum progesterone appeared to be the single most specific biomarker for distinguishing viable from nonviable pregnancies. When a dual-biomarker strategy was applied, combining serum progesterone with hCG, specificity improved significantly, which suggests that a multiple biomarker strategy might help distinguish viable from nonviable pregnancies in early gestation.