Serum progesterone in predicting pregnancy outcome after assisted reproductive technology.
PURPOSE: Our purpose was to determine whether serum progesterone predicts pregnancy outcome after superovulation. METHODS: One hundred twenty-three consecutively pregnant patients were divided into three groups: group I, 55 patients following superovulation for assisted reproductive technologies; group II, 23 patients after correction of oligoovulation; and group III, 45 patients who conceived spontaneously. When beta-human chorionic gonadotropin was positive, progesterone was measured on the same serum sample. A serum progesterone level of 45 microns/L was set to differentiate between nonviable pregnancy and ongoing pregnancy. RESULTS: In group I, zero (0%) of 38 ongoing pregnancies and 10 (59%) of 17 nonviable pregnancies were observed with a progesterone level of < 45 microns/L [14.2 ng/ml (P < 0.001)]. In group II, 4 (27%) of 15 ongoing pregnancies and 5 (63%) of 8 nonviable pregnancies had a progesterone level of < 45 microns/L (P = NS). In group III, 10 (42%) of 24 ongoing pregnancies and 15 (71%) of 21 nonviable pregnancies were observed with a progesterone level of < 45 microns/L (14.2 ng/ml) (P = NS). CONCLUSIONS: A serum progesterone level of < 45 nM predicts nonviable pregnancy after superovulation for assisted reproductive technology.